ProtAttn-QuadNet: An attention-based deep learning framework for protein–protein interaction prediction using ProtBERT embeddings

Md. Shahidul Islam; Md. Muhtasim Rahman Mim; Md. Raihan Kabir

doi:10.1371/journal.pone.0349433

Peer Review History

Original SubmissionNovember 26, 2025
26 Jan 2026 Decision Letter - Musa Aydin, Editor -->PONE-D-25-63548-->-->ProtAttn-QuadNet: An attention-based deep learning framework for protein–protein interaction prediction using ProtBERT embeddings-->-->PLOS One Dear Dr. Islam, 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 Mar 12 2026 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 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: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols. We look forward to receiving your revised manuscript. Kind regards, Musa Aydin, 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. 2. If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise. [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 Reviewer #2: Partly ******** -->2. Has the statistical analysis been performed appropriately and rigorously? --> Reviewer #1: Yes Reviewer #2: 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 Reviewer #2: 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 Reviewer #2: 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: Page 7, Abstract: Align accuracy and AUC values with those in the Results section. Page 10, Table 1: Consider adding standard deviations if multiple runs were performed. Page 13, Data Preprocessing: Clarify why robust scaling was chosen over other methods (e.g., MinMax, StandardScaler) in more detail. Page 14, Feature Engineering: Briefly justify the noise injection (Eq. 8) and its impact on generalization. References: Ensure all in-text citations are in the reference list and formatted consistently. Reviewer #2: • The model presents an innovative approach to protein representation for PPI tasks, utilizing complementary methods to capture both similarity and difference information between protein pairs. • The extracted representations are processed through attention mechanisms in multiple configurations, then trained within a multi-task learning framework that derives information from binary interaction labels. • Figures are not properly integrated into the text and instead appear individually on separate pages following the references. • Section ordering should follow standard convention: Results and Discussion must appear after Materials and Methods. • While evaluation metrics can be briefly mentioned as standard, their definitions need not be elaborated. • Table 1 values should not be redundantly listed in the text. Commentary on the achieved metrics is appropriate, but verbatim repetition of tabular data is unnecessary. • There is a critical problem with comparison to other models. It is stated in the article verbatim “Across all evaluated datasets, ProtAttn-QuadNet demonstrated consistently superior predictive performance”. This statement is not proven at all. Only given result for ProtAttn-QuadNet model is on UniProt dataset. But no other model is evaluated on this dataset, making a direct comparison impossible. It is also stated “Notably, even on smaller organism-specific datasets, such as S. cerevisiae and H. pylori, ProtAttn-QuadNet either matched or exceeded the highest reported F1-scores from previous studies, including xCAPT5 and SCMPPI.”. Results achieved by ProtAttn-QuadNet in these datasets are not shared anywhere in the article. Authors must share these results in order to support this claim. • Figure 6 contains multiple discrepancies with both the textual description and implementation. Minor issues include omission of projection layers and inconsistent naming (only x_diff transforms to h_diff while other inputs retain their names). More substantially, the attention mechanism connections are unclear—all inputs appear to interact directly between the attention and hidden layers, which contradicts both the text and code. The hidden layer following the attention layer is ambiguous: if this represents the FFN applied after MultiHeadAttention, why are these components not grouped into an AttentionBlock as described in the text? Given its central importance, this figure requires careful revision. • Notation for A is used inconsistently. Initially, A denotes the output of MultiHeadAttention used in computing h. Subsequently, A represents the second output of AttentionBlock (the attention weights in the code, which remain unused). This creates ambiguity suggesting the MultiHeadAttention output is split for separate use. These should be labeled differently. Furthermore, the purpose of extracting attention weights is unclear given they are utilized neither in the code nor discussed in the manuscript. • Tasks beyond interaction prediction require more detailed specification, particularly their exact computational formulation. For example, binding strength is computed in the code as: interaction_conf = torch.abs(predictions['interaction'].squeeze() - 0.5) * 2; binding_target = interaction_conf * targets.float(). Such calculations should be explicitly presented and justified in the methodology. • Axis labels and plot titles should be consistent between figures 1 and 2. • Bias notation alternates between bold and italic; one convention should be used throughout. ******** -->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: Yes: Ghassan Abdul-Majeed 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.] To ensure your figures meet our technical requirements, please review our figure guidelines: https://journals.plos.org/plosone/s/figures You may also use PLOS’s free figure tool, NAAS, to help you prepare publication quality figures: https://journals.plos.org/plosone/s/figures#loc-tools-for-figure-preparation. NAAS will assess whether your figures meet our technical requirements by comparing each figure against our figure specifications. https://doi.org/10.1371/journal.pone.0349433.r001
Revision 1
24 Mar 2026 Author Response Reviewer #1 Thank you for your careful review and valuable feedback. We sincerely appreciate your time and insightful suggestions. We have addressed all your comments and revised the manuscript accordingly. Please note that the manuscript has been restructured to comply with the journal guidelines; consequently, the page and equation numbering may differ in the revised version. Page 7, Abstract: Align accuracy and AUC values with those in the Results section. Response: Thank you very much for your thoughtful comments. We have carefully rechecked the reported values and confirm that they are correct for the independent test set. However, to avoid any potential confusion between the test and validation sets, we have revised the corresponding text in the manuscript as follows: “Evaluated on the independent test set of a large-scale dataset from UniProt, ProtAttn-QuadNet achieves 97.16% accuracy (AUC-ROC 99.00%) on balanced data and 99.19% accuracy (AUC-ROC 99.76%) on oversampled datasets, surpassing several recent state-of-the-art PPI prediction methods. ” Page 10, Table 1: Consider adding standard deviations if multiple runs were performed. Response: Thank you for your comment. We did not perform a statistical deviation test; however, we will consider including it in the final version. Page 13, Data Preprocessing: Clarify why robust scaling was chosen over other methods (e.g., MinMax, StandardScaler) in more detail. Response: Thank you for the helpful suggestion. We have expanded the explanation in the manuscript to clarify why RobustScaler was selected over StandardScaler and MinMaxScaler, emphasizing that protein embeddings may exhibit non-Gaussian distributions and outliers, making IQR–based scaling more stable for such features. Please see the highlighted verison of the manuscript (page 3, Data Processing). Page 14, Feature Engineering: Briefly justify the noise injection (Eq. 8) and its impact on generalization. Response: We have added a new paragraph titled “Gaussian Noise Injection for Data Augmentation” before Eq. 3 (previously Eq. 8) to clarify the rationale for noise augmentation. The paragraph explains that small Gaussian perturbations to ProtBert embeddings improve model robustness and generalization by encouraging the model to learn stable interaction patterns rather than memorizing exact embedding values. Please see page 4 in the manuscript. References: Ensure all in-text citations are in the reference list and formatted consistently. Response: Thank you for your comment. We have carefully reviewed the reference list and confirm that all references are properly cited within the text and are formatted appropriately. Reviewer #2 We sincerely thank you for your thorough review and insightful feedback. We greatly appreciate the time and effort you invested in evaluating our manuscript. We have carefully addressed all your comments and have updated the manuscript accordingly. • The model presents an innovative approach to protein representation for PPI tasks, utilizing complementary methods to capture both similarity and difference information between protein pairs. Response: Thank you very much for your comment. • The extracted representations are processed through attention mechanisms in multiple configurations, then trained within a multi-task learning framework that derives information from binary interaction labels. Response: Thank you. We appreciate your feedback. • Figures are not properly integrated into the text and instead appear individually on separate pages following the references. Response: Thank you very much for your valuable comment. However, we have followed the PLOS ONE submission guidelines, which state: “Do not include figures in the main manuscript file. Each figure must be prepared and submitted as an individual file.” Therefore, we have prepared and submitted the figures separately in accordance with the journal’s requirements. • Section ordering should follow standard convention: Results and Discussion must appear after Materials and Methods. Response: Thank you very much for your careful observation. We have revised the manuscript to comply with the PLOS ONE submission guidelines. The Materials and Methods section has now been placed before the Results and Discussion section as required. • While evaluation metrics can be briefly mentioned as standard, their definitions need not be elaborated. Response: As suggested, we have removed the detailed formulas and now briefly describe the evaluation metrics. • Table 1 values should not be redundantly listed in the text. Commentary on the achieved metrics is appropriate, but verbatim repetition of tabular data is unnecessary. Response: Thank you for your valuable comment. We have revised the text to avoid verbatim repetition of the values in Table 1. Instead, we now provide a concise commentary on the model’s performance, highlighting trends and key observations while referring to the table for detailed metrics. Please see page 9 in the manuscript. • There is a critical problem with comparison to other models. It is stated in the article verbatim “Across all evaluated datasets, ProtAttn-QuadNet demonstrated consistently superior predictive performance”. This statement is not proven at all. Only given result for ProtAttn-QuadNet model is on UniProt dataset. But no other model is evaluated on this dataset, making a direct comparison impossible. It is also stated “Notably, even on smaller organism-specific datasets, such as S. cerevisiae and H. pylori, ProtAttn-QuadNet either matched or exceeded the highest reported F1-scores from previous studies, including xCAPT5 and SCMPPI.”. Results achieved by ProtAttn-QuadNet in these datasets are not shared anywhere in the article. Authors must share these results in order to support this claim. Response: Thank you so much. We have clarified in the revised manuscript that comparisons with previous studies are based on overlapping organisms rather than identical datasets. Please see the revised text in the manuscript. • Figure 6 contains multiple discrepancies with both the textual description and implementation. Minor issues include omission of projection layers and inconsistent naming (only x_diff transforms to h_diff while other inputs retain their names). More substantially, the attention mechanism connections are unclear—all inputs appear to interact directly between the attention and hidden layers, which contradicts both the text and code. The hidden layer following the attention layer is ambiguous: if this represents the FFN applied after MultiHeadAttention, why are these components not grouped into an AttentionBlock as described in the text? Given its central importance, this figure requires careful revision. Response: Thank you so much for your critical observation. We have now updated the figure according to the text to remove any confusion. Which also removed the connection issues. • Notation for A is used inconsistently. Initially, A denotes the output of MultiHeadAttention used in computing h. Subsequently, A represents the second output of AttentionBlock (the attention weights in the code, which remain unused). This creates ambiguity suggesting the MultiHeadAttention output is split for separate use. These should be labeled differently. Furthermore, the purpose of extracting attention weights is unclear given they are utilized neither in the code nor discussed in the manuscript. Response: We sincerely thank you for carefully identifying this issue, which has helped us improve the consistency of the manuscript. Each AttentionBlock returns the transformed feature representation (h) and the corresponding attention weights (A). We used only the feature representations (h) for downstream prediction. The attention weights are present in the implementation for potential analysis; however, they are neither captured during the function call nor utilized in subsequent computations. Therefore, references to the attention weights have been removed from the manuscript, and the notation A now consistently refers only to the output of MultiHeadAttention. • Tasks beyond interaction prediction require more detailed specification, particularly their exact computational formulation. For example, binding strength is computed in the code as: interaction_conf = torch.abs(predictions['interaction'].squeeze() - 0.5) * 2; binding_target = interaction_conf * targets.float(). Such calculations should be explicitly presented and justified in the methodology. Response: Thank you for your comment. We have now discussed the loss function in details including its exact formulation. Please see equations 13 to 18 in the manuscript (page 7). • Axis labels and plot titles should be consistent between figures 1 and 2. Response: Thank you for the comment. We have now made both figures 1 and 2 consistent in title and axis labels. • Bias notation alternates between bold and italic; one convention should be used throughout. Response: Thank you for your careful observation. We have revised the manuscript to ensure consistent notation throughout. Specifically, all bias terms are now presented in bold font, as they represent vectors. Attachments Attachment Submitted filename: Response to Reviewers Comments - ProtAttn-QuadNet.docx https://doi.org/10.1371/journal.pone.0349433.r002
1 May 2026 Decision Letter - Musa Aydin, Editor ProtAttn-QuadNet: An attention-based deep learning framework for protein–protein interaction prediction using ProtBERT embeddings PONE-D-25-63548R1 Dear Dr. Islam, 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 will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. For questions related to billing, please contact billing support. 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, Musa Aydin, 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 #1: All comments have been addressed 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 #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: Dear Authors, Thank you for fixing all my comments. The revised manuscript is now much better than the first one. 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 #1: Yes: Ghassan Abdul-Majeed Reviewer #2: No ******** https://doi.org/10.1371/journal.pone.0349433.r003
Formally Accepted
Acceptance Letter - Musa Aydin, Editor PONE-D-25-63548R1 PLOS One Dear Dr. Islam, I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS One. Congratulations! Your manuscript is now being handed over to our production team. At this stage, our production department will prepare your paper for publication. This includes ensuring the following: * All references, tables, and figures are properly cited * All relevant supporting information is included in the manuscript submission, * There are no issues that prevent the paper from being properly typeset You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps. Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. 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. You will receive an invoice from PLOS for your publication fee after your manuscript has reached the completed accept phase. If you receive an email requesting payment before acceptance or for any other service, this may be a phishing scheme. Learn how to identify phishing emails and protect your accounts at https://explore.plos.org/phishing. If we can help with anything else, please email us at customercare@plos.org. Thank you for submitting your work to PLOS ONE and supporting open access. Kind regards, PLOS ONE Editorial Office Staff on behalf of Assoc. Prof. Musa Aydin Academic Editor PLOS One https://doi.org/10.1371/journal.pone.0349433.r004

Open letter on the publication of peer review reports

PLOS recognizes the benefits of transparency in the peer review process. Therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. Reviewers remain anonymous, unless they choose to reveal their names.

We encourage other journals to join us in this initiative. We hope that our action inspires the community, including researchers, research funders, and research institutions, to recognize the benefits of published peer review reports for all parts of the research system.

Learn more at ASAPbio .