Utilizing CNNs for classification and uncertainty quantification for 15 families of European fly pollinators

Thomas Stark; Michael Wurm; Valentin Ştefan; Felicitas Wolf; Hannes Taubenböck; Tiffany M. Knight

doi:10.1371/journal.pone.0323984

Peer Review History

Original SubmissionApril 20, 2025
2 Jun 2025 Decision Letter - Ramzi Mansour, Editor PONE-D-25-20622Utilizing CNNs for classification and uncertainty quantification for 15 families of European fly pollinatorsPLOS ONE Dear Dr. Stark, 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 Jul 17 2025 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: 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, Ramzi Mansour 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. Please provide a complete Data Availability Statement in the submission form, ensuring you include all necessary access information or a reason for why you are unable to make your data freely accessible. If your research concerns only data provided within your submission, please write "All data are in the manuscript and/or supporting information files" as your Data Availability Statement. [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: I Don't Know ****** 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: This is a very well written and presented manuscript; discussion and conclusions are well thought through, and relevant literature used. It will be of interest to many others regarding methods and overall conclusions. Minor issues • Line 168 – how did the bounding box but also the preprocessing get done, what software environment? • Fig3 misleading legend, should say "With" and "Without" (rather than False/True), the figure can be interpreted differently with the words False/True • Fig 3 caption – spelling “boundinx", should be Bounding • Abstract states "Our dataset comprises a wide range of morphological and phylogenetic features, such as wing venation patterns and wing shapes" but this is misleading; the images used were general habitus images of the body. Yes "features of the wings" are part of this general body but they are not the only feature. The statement in the abstract implies wing features are the central part of the images. • Early discussion (lines 305 +) regarding the time to make bounding boxes; unsure if the authors know of Roboflow (www.roboflow.com) - it has some semi automated features and is very easy to use. One thing the authors could examine in any future work is the ranges and overlap of confidence scores of those images that are scored correctly vs incorrectly. There is some discussion of this on lines around 380-390. This connects with "trustworthiness" that the authors mention in the introduction. It is an issue that we are facing in our lab group - where is there overlap in scores between right and wrong, and so it is difficult for the user to be confident of the score. Reviewer #2: General comments I enjoyed reading this study and there were some interesting methodology and results however some information was lacking. Further details are required in the materials and methods for transparency and the figure captions need fuller explanations so that if viewed independently the reader understands them. A key aspect of this study are the uncertainty quantification and mean confidence calculations; these provide a potentially useful metric for researchers using machine learning classifiers however some methodological details are missing. There is also some imbalance in the number of images for each of the families and I would be interested to see the f1 scores to determine whether this has impacted CNN performance. If these issues, further detailed in the comments below were addressed I would recommend this study for publication. Abstract I would like more information about the methodology here and some more results – e.g. what method did you use to determine accuracy and prediction confidence? What was the improvement in accuracy using bounding boxes? Introduction Line 12: A definition of a CNN would be useful here. Line 28: Further expansion of this paragraph would help to highlight the rationale for your study Line 36: Clarification of uncertainty estimation in relation to previous works and how it appliesto this study would be beneficial in this paragraph. Line 58: a brief description of your uncertainty quantification methods in this paragraph would emphasise the novelty of your approach Materials and Methods General - Further information about Diptera and their identification features would be interesting, particularly examples of wing venation. Including the formulae for the mean confidence score would provide further clarity. Line 79: I think this paragraph would be better placed in the introduction Line 147: Which image augmentations were used and in which proportions? Line 154: What percentage of dropout did you use? Line 160/161: Did you use augmented data when you trained these models? Figure 1: More detail in the caption is needed – are the confidence figures for all species? Line 197: This paragraph is a little ambiguous, which configurations did you investigate? Learning rate, batch size etc? Experimental set up: Information regarding the performance metrics and how they are calculated is missing. The Kappa score is calculated but not explained. Results Line 217: I think you are understating your results a little here, all your models achieved OAs in excess of 88% and mean confidence of more than 78%. The Kappa score could also be mentioned here. Line 232 and 235: You could add the mean improvement of OA and mean confidence across all CNNs for further clarity. Figure 2: I would include the percentages only in the confusion matrices to make the figure more readable The explanation of figure 5 is missing in this section. Discussion Line 309: The sentence “The labor-intensive nature of this manual approach underscores the potential benefits of automation” seems to contradict the contribution of your work – I think a more interesting issue is the trade-off between the labour required to crop the images and the improvement in OA and mean confidence scores. Figure 5: This figure is somewhat confusing – are the green boxes examples of two different species from the same family that have been correctly predicted? Further information in the legend would clarify this. Confusion between Diptera families: In the second and third paragraphs, the two families you are comparing Fanniidae and Muscidae have 521 and 2163 images respectively, what impact does this class imbalance between have on their OA and mean confidence? ****** 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: Darren Ward 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. https://doi.org/10.1371/journal.pone.0323984.r001
Revision 1
23 Jul 2025 Author Response Response to the reviewers Title: Utilizing CNNs for classification and uncertainty quantification for 15 families of European fly pollinators Manuscript Number: PONE-D-25-20622 Dear Irene Nathalie Fernandez Tolentino, Dear Editors, Thank you very much for considering our manuscript for publication PLOS ONE subject to the requested revisions based on reviewers’ comments. Please find below a point-by-point answer to the comments (bold, cursive and bullet points) in blue and new text which was added to the main manuscript in red. We acknowledge the valuable reviewers’ comments and hope that the revised manuscript meets all specifications with respect to the changes made. Sincerely, Thomas Stark Reviewer #1: This is a very well written and presented manuscript; discussion and conclusions are well thought through, and relevant literature used. It will be of interest to many others regarding methods and overall conclusions. • (Line 168) How did the bounding box but also the preprocessing get done, what software environment? Thank you for bringing this to our attention we added more detail of making the bounding boxes and expanded the preprocessing process in more detail. The bounding box generation was conducted in a two-step process. First, we used a pre-trained object detection model based on the approach described in [23] to automatically generate initial bounding boxes around the Diptera specimens. In a second step, these boxes were manually refined by two expert ecologists to ensure high annotation accuracy. This manual correction was performed for all ~30,000 images using the open-source tool LabelImg within a Python-based environment. This combination of automated detection and expert validation ensured both efficiency and high-quality ground truth for subsequent model training. (L.213-218) In Figure 1 our methodological approach is illustrated in detail. Initially, bounding boxes are created for 29,374 images to crop them, ensuring that only the entire body of the Diptera is visible while removing any background as seen in Figure 1(A). Bounding Boxes were first generated using a pre-trained object detection model from [23] and then manually corrected by two expert ecologists using the open-source tool LabelImg in a Python environment. We are not entirely sure whether the preprocessing refers to image preprocessing or data selection preprocessing. Nevertheless, for image preprocessing, we applied z-score normalization using the RGB mean and standard deviation values from ImageNet, which we also used for pretraining our three CNN models. This normalization helps standardize pixel distributions across images and is commonly used in deep learning pipelines to improve model convergence and performance [32,37,38] Regarding dataset curation, this was a more involved step due to the high variability of GBIF images. We invested considerable effort into selecting images that depict adult Diptera on flowers or in natural environments, while excluding larval stages, museum specimens, or highly artificial backgrounds. While GBIF image tags were partially helpful for filtering, most of this refinement took place during the bounding box creation and correction process, where we manually removed unsuitable samples to ensure a consistent and high-quality dataset. • Fig3 misleading legend, should say "With" and "Without" (rather than False/True), the figure can be interpreted differently with the words False/True Thank you for highlighting this typo, we fixed it in the revised manuscript. • Fig 3 caption – spelling “boundinx", should be Bounding Thank you again for highlighting this typo, we fixed it in the revised manuscript. Fig 3. Confidence values for EfficientNetB4 for all 15 Diptera families for correct classification results with and without cropping its Bounding Boxes. • The abstract states "Our dataset comprises a wide range of morphological and phylogenetic features, such as wing venation patterns and wing shapes" but this is misleading; the images used were general habitus images of the body. Yes "features of the wings" are part of this general body but they are not the only feature. The statement in the abstract implies wing features are the central part of the images. Thank you for highlighting this point. We agree that the original phrasing may have been too strong and will soften the wording in the abstract accordingly. Pollination is essential for maintaining biodiversity and ensuring food security, and in Europe it is primarily mediated by four insect orders (Coleoptera, Diptera, Hymenoptera, Lepidoptera). However, traditional monitoring methods are costly and time consuming. Although recent automation efforts have focused on butterflies and bees, flies, a diverse and ecologically important group of pollinators, have received comparatively little attention, likely due to the challenges posed by their subtle morphological differences. In this study, we investigate the application of Convolutional Neural Networks (CNNs) for classifying 15 European pollinating fly families and quantifying the associated classification uncertainty. In curating our dataset, we ensured that the images of Diptera captured diverse visual characteristics relevant for classification, including wing morphology and general body habitus. We evaluated the performance of three CNNs, ResNet18, MobileNetV3, and EfficientNetB4 and estimated the prediction confidence using Monte Carlo methods, combining test-time augmentation and dropout to approximate both aleatoric and epistemic uncertainty. We demonstrate the effectiveness of these models in accurately distinguishing fly families. We achieved an overall accuracy of up to 95.61%, with a mean relative increase in accuracy of 5.58% when comparing uncropped to cropped images. Furthermore, cropping images to the Diptera bounding boxes not only improved classification performance across all models but also increased mean prediction confidence by 8.56%, effectively reducing misclassifications among families. This approach represents a significant advance in automated pollinator monitoring and has promising implications for both scientific research and practical applications. • Early discussion (lines 305 +) regarding the time to make bounding boxes; unsure if the authors know of Roboflow (www.roboflow.com) - it has some semi automated features and is very easy to use. Thank you for this helpful suggestion. We were only loosely aware of Roboflow but had not used it in our workflow. Its semi-automated features indeed seem very helpful, and we now mention in the discussion that such tools could support more efficient annotation in future work. (L.375-388) Creating bounding boxes manually is an extremely time-consuming process. For the 29,374 images used in this study, bounding boxes had to be created by experts, following a consistent workflow to ensure that each specimen is cropped uniformly. This task requires not only significant time but also specialized expertise to accurately delineate the objects of interest. The labor-intensive nature of this manual approach underscores the potential benefits of automation. Also semi-automated labeling tools, could further support and streamline this process. It would be very interesting to explore the use of object detection algorithms to automatically crop images similar to the methods used in [49] and [25]. • One thing the authors could examine in any future work is the ranges and overlap of confidence scores of those images that are scored correctly vs incorrectly. There is some discussion of this on lines around 380-390. This connects with "trustworthiness" that the authors mention in the introduction. It is an issue that we are facing in our lab group - where is there overlap in scores between right and wrong, and so it is difficult for the user to be confident of the score. Thank you for this very interesting suggestion. We have also noticed similar issues in our results and agree that analyzing the overlap in confidence scores between correctly and incorrectly classified images is a promising direction. This is definitely a topic we will look into more closely in future work, especially now knowing that other groups are encountering a similar challenge. Reviewer #2: I enjoyed reading this study and there were some interesting methodology and results however some information was lacking. Further details are required in the materials and methods for transparency and the figure captions need fuller explanations so that if viewed independently the reader understands them. A key aspect of this study are the uncertainty quantification and mean confidence calculations; these provide a potentially useful metric for researchers using machine learning classifiers however some methodological details are missing. • There is also some imbalance in the number of images for each of the families and I would be interested to see the f1 scores to determine whether this has impacted CNN performance. Thank you for your valuable observation regarding potential class imbalance and the relevance of F1 scores. While we did observe some variation in the number of images per fly family, we do not consider the imbalance to be substantial enough to significantly skew model performance. To evaluate this further, we did compute the macro-averaged F1 scores alongside other metrics. However, we chose to focus on overall accuracy and Cohen’s kappa in the manuscript to maintain clarity and avoid redundancy, as the F1 scores were very similar to these primary metrics. Nevertheless, we appreciate your suggestion and have now included a table comparing accuracy, kappa, and F1 score per model and variant. As shown in the table below, the F1 scores follow the same trends as the accuracy and kappa metrics. For example, the EfficientNet model with bounding box cropping (BB) achieved the highest values across all three metrics (Accuracy: 95.61%, Kappa: 0.9590, F1: 0.9622), while the noBB variants consistently performed lower. These results support our interpretation that minor class imbalance did not substantially impact model performance. Model Variant Accuracy Kappa F1 MobileNet noBB 88.78% 0.8887 0.8963 ResNet noBB 88.58% 0.8895 0.8973 EfficientNet noBB 90.35% 0.9044 0.9110 MobileNet BB 93.87% 0.9401 0.9444 ResNet BB 93.16% 0.9384 0.9426 EfficientNet BB 95.61% 0.9590 0.9622 • Abstract: I would like more information about the methodology here and some more results – e.g. what method did you use to determine accuracy and prediction confidence? What was the improvement in accuracy using bounding boxes? Thank you for highlighting this important point. We have revised the abstract to include more specific details regarding our methodology and key results. In particular, we now explicitly state that overall accuracy (OA) was used as the main performance metric, and that prediction confidence and uncertainty were quantified using Monte Carlo methods, combining test-time augmentation and dropout to estimate both aleatoric and epistemic uncertainty. Additionally, we clarified the performance gains achieved through image cropping. Specifically, we now report that EfficientNetB4 reached an overall accuracy of 95.61%, representing a relative improvement of approximately 5.82% over the uncropped version. We also mention that, on average, cropping images to the bounding boxes of Diptera resulted in a 5.58% increase in classification accuracy and an 8.56% increase in prediction confidence across all models. These additions help emphasize the benefit of expert-defined cropping in boosting model reliability. We hope these clarifications improve the completeness and transparency of the abstract. Pollination is essential for maintaining biodiversity and ensuring food security, and in Europe it is primarily mediated by four insect orders (Coleoptera, Diptera, Hymenoptera, Lepidoptera). However, traditional monitoring methods are costly and time consuming. Although recent automation efforts have focused on butterflies and bees, flies, a diverse and ecologically important group of pollinators, have received comparatively little attention, likely due to the challenges posed by their subtle morphological differences. In this study, we investigate the application of Convolutional Neural Networks (CNNs) for classifying 15 European pollinating fly families and quantifying the associated classification uncertainty. In curating our dataset, we ensured that the images of Diptera captured diverse visual characteristics relevant for classification, including wing morphology and general body habitus. We evaluated the performance of three CNNs, ResNet18, MobileNetV3, and EfficientNetB4 and estimated the prediction confidence using Monte Carlo methods, combining test-time augmentation and dropout to approximate both aleatoric and epistemic uncertainty. We demonstrate the effectiveness of these models in accurately distinguishing fly families. We achieved an overall accuracy of up to 95.61%, with a mean relative increase in accuracy of 5.58% when comparing uncropped to cropped images. Furthermore, cropping images to the Diptera bounding boxes not only improved classification performance across all models but also increased mean prediction confidence by 8.56%, effectively reducing misclassifications among families. This approach represents a significant advance in automated pollinator monitoring and has promising implications for both scientific research and practical applications. Introduction • Line 12: A definition of a CNN would be useful here. We have added a brief definition of Convolutional Neural Networks (CNNs) directly after their first mention in the text to improve clarity for readers unfamiliar with the concept. Additionally, we now include two standard references to foundational CNN [11, 12] to support this explanation. (L.10-15) The limited availability of images for flies, compared to more charismatic taxa of pollinators, makes it difficult to create a comprehensive dataset, which is essential for successfully training Convolutional Neural Networks (CNNs) [9, 10]. CNNs are a class of deep learning models particularly well-suited for image recognition tasks due to their ability to automatically learn hierarchical spatial features from raw pixels [11, 12]. • Line 28: Further expansion of this paragraph would help to highlight the rationale for your study Thank you very much for highlighting the importance of clearly positioning our study within the broader context of deep learning applications for arthropod classification. In response, we have revised and expanded the background section to better contextualize our work and explain its rationale. Specifically, we now provide a more detailed overview of how deep learning has been applied to various arthropod groups, citing several recent studies across different taxa. We also added a dedicated paragraph emphasizing that pollinating flies—despite being the world’s second most important group of pollinators after bees—have received comparatively little attention in this research area. To support this point, we included several key references [6,27,28,29] demonstrating the ecological and functional relevance of Diptera, especially in regions or conditions where bees are less prevalent. These additions strengthen the motivation for our study and highlight the gap our work aims to address: the need for robust deep learning models capable of identifying pollinating flies, at least to the family level, for comprehensive biodiversity monitoring and conservation efforts. (L.32-41) However, pollinating flies have not yet been systematically addressed in this line of research, despite being the world’s second most important and abundant pollinators after bees [6,27]. Flies, particularly syrphids and other Diptera, make vital contributions to both natural ecosystems an Attachments Attachment Submitted filename: Letter_to_reviewers_PONE-D-25-20622.pdf https://doi.org/10.1371/journal.pone.0323984.r002
14 Aug 2025 Decision Letter - Ramzi Mansour, Editor Utilizing CNNs for classification and uncertainty quantification for 15 families of European fly pollinators PONE-D-25-20622R1 Dear Dr. Stark, 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, Ramzi Mansour 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: (No Response) Reviewer #2: I would like to thank the author for addressing my comments comprehensively and appreciate the improvements made to your article. ****** 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 ******** https://doi.org/10.1371/journal.pone.0323984.r003
Formally Accepted
Acceptance Letter - Ramzi Mansour, Editor PONE-D-25-20622R1 PLOS ONE Dear Dr. Stark, 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 Dr. Ramzi Mansour Academic Editor PLOS ONE https://doi.org/10.1371/journal.pone.0323984.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 .