An exploratory machine learning study on paediatric abdominal pain phenotyping and prediction

Kazuya Takahashi; Michalina Lubiatowska; Huma Shehwana; James K. Ruffle; John A. Williams; Animesh Acharjee; Shuji Terai; Georgios V. Gkoutos; Humayoon Satti; Qasim Aziz

doi:10.1371/journal.pone.0336215

Peer Review History

Original SubmissionJuly 4, 2025
18 Sep 2025 Decision Letter - Hany Abo-Haded, Editor PLOS ONE Dear Dr. Takahashi, 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. ACADEMIC EDITOR COMMENTS: Please submit your revised manuscript by Nov 02 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. 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, Hany Mahmoud Abo-Haded, MD 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 note that PLOS One has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, we expect all author-generated code to be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/plosone/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse. 3. Thank you for stating the following financial disclosure: [This study was supported by JA Niigata Kouseiren Grant (Niigata University School of Medicine) and JSPS Grants-in-Aid for Scientific Research (grant number: 22K16013).]. Please state what role the funders took in the study. If the funders had no role, please state: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript." If this statement is not correct you must amend it as needed. Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf. 4. 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. 5. 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. Additional Editor Comments: Reviewer #1: Reviewer #2: [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? Reviewer #1: Partly Reviewer #2: Yes ******** 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 Reviewer #1: Yes Reviewer #2: Yes ****** 4. Is the manuscript presented in an intelligible fashion and written in standard English??> Reviewer #1: Yes Reviewer #2: Yes ****** Reviewer #1: The manuscript was a great read, introducing AI and Abdominal Pain (AP). AI and medical associations are the wave of the future but still needs perfecting. In agreeance with AI facilitating a deep dive into the human diagnosis. It can be difficult in choosing the supervised and unsupervised learning as they are both beneficial in human exploration. Unsupervised learning needs the human expertise with the outcomes. There are area of this manuscript that need details on technique and methods used, as AI is in its introductory phase. The association is a bit farfetched or novel. Major Revision/Recommendations: 1. Supervised vs unsupervised learning: From a clinical aspect, unsupervised was the correct step but why not complete both methods. You identified in your manuscript that unsupervised learning was implemented due to exploratory features, I agree. This should be explained or made clear. Also, supervised could be very beneficial, in predicting future outcomes. 2. Illustrations should be added using cluster analysis, hierarchical structure of the identified clusters, etc. Emerging topics should be visual, so that your audience receives your point of view. 3. The cluster were mainly focused on allergic disease which is not a common in abdominal pain. I would explain this in detail, mentioning this finding from beginning to end. Are you able to explain the algorithm or coding. The result of allergic disease associated with abdominal pain is a bit farfetched with their being more common associations. 4. The significance of the population being mostly Pakistan and White could be multifactorial (environmental, etc.) 5. The phenotypes seemed unmatched with actual societal groups with abdominal pain or maybe novel new finding. 6. If AI is main objective, this is great but if using the association to come to a conclusion more work is needed. If this is a manuscript on allergic disease, AI and maternal comorbidities, great read but to associate with Abdominal Pain is a bit farfetched. 7. Your audience needs to be taught ad convinced that AI is appropriate for science and will generate conclusive results. Reviewer #2: The study by Kazuya Takahashi et al. explores pediatric abdominal pain (AP) using machine learning (ML) to identify phenotypes and predict risk factors. Analyzing data from 13,790 children in the Born in Bradford cohort, the researchers identified three AP phenotypes: allergic predisposition, maternal comorbidities, and minimal comorbidities. Allergic diseases and maternal health issues significantly increased the frequency of AP, with 17.6% of children having ≥3 allergic diseases and 25.6% of children with ≥3 maternal comorbidities experiencing AP. A supervised ML model achieved moderate predictive performance (AUC 0.67), highlighting ethnicity, pediatric allergic diseases, and maternal comorbidities as key factors. Risk stratification showed AP rates ranging from 18.9% (<40% probability) to 100% (>60% probability). These findings emphasize the role of genetic, environmental, and maternal influences in the development of AP, offering insights for future research and personalized interventions. The study question is relevant to clinical practice, and the authors conducted a thorough literature review. The analysis of data from 13,790 children provides robust statistical power and generalizability. The major findings are clearly presented, and the research objectives are effectively addressed. The results support the conclusions, and the manuscript is well-written. However, the manuscript could be improved by addressing the following issues: 1. The authors should discuss in greater detail the gap in existing knowledge regarding AP phenotypes and predictive factors to better justify the need for the study. 2. A brief overview of the methods (e.g., ML clustering and predictive modeling) would help readers understand how the study addresses the research gap. 3. Summarizing the most important results at the beginning of each subsection in the Results section would help readers quickly grasp the main takeaways. 4. The Discussion section could be strengthened by incorporating more data from previous studies to contextualize the findings and highlight how this study advances knowledge on pediatric abdominal pain. 5. The authors should provide deeper insights into the clinical implications of the identified phenotypes, including how they could inform personalized treatment strategies. 6. More specific recommendations for future studies would be beneficial, such as incorporating pathophysiological data, validating findings in diverse populations, and exploring genetic factors. ****** 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: Nicole Y Fatheree 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.0336215.r001
Revision 1
14 Oct 2025 Author Response 13th October 2025 Emily Chenette Editor-In-Chief PLOS ONE Dear Editor, Re: An Exploratory Machine Learning Study on Paediatric Abdominal Pain Phenotyping and Prediction (PONE-D-25-34689) Thank you for your decision dated 18th September 2025, enclosing the reviewers' comments. We have carefully reviewed the comments and have revised the manuscript accordingly. Our responses are given in a point-by-point manner below. Modifications to the manuscript are written in red. We hope the revised manuscript is now suitable for publication and look forward to hearing from you in due course. Yours sincerely, Kazuya Takahashi Centre for Neuroscience, Surgery and Trauma, Wingate Institute of Neurogastroenterology, Blizard Institute, Barts and the London School of Medicine and Dentistry, Queen Mary University of London Email: kazuya911@med.niigata-u.ac.jp Comments from the academic editor Please complete the explanation of both learning methods (the supervised and the unsupervised ) Response: Thank you for the valuable feedback and the opportunity to further clarify our methodology. We have expanded the description of both supervised and unsupervised machine learning approaches to provide a clearer understanding of their principles and applications in our study (page 9, line 17 to page 13, line 8). We hope that the revised explanation enhances the readers’ comprehension of our methods. Comments from Reviewer 1 The manuscript was a great read, introducing AI and Abdominal Pain (AP). AI and medical associations are the wave of the future but still needs perfecting. In agreeance with AI facilitating a deep dive into the human diagnosis. It can be difficult in choosing the supervised and unsupervised learning as they are both beneficial in human exploration. Unsupervised learning needs the human expertise with the outcomes. There are area of this manuscript that need details on technique and methods used, as AI is in its introductory phase. The association is a bit farfetched or novel. Major Revision/Recommendations: 1. Supervised vs unsupervised learning: From a clinical aspect, unsupervised was the correct step but why not complete both methods. You identified in your manuscript that unsupervised learning was implemented due to exploratory features, I agree. This should be explained or made clear. Also, supervised could be very beneficial, in predicting future outcomes. Response: Thank you for providing us with the opportunity to elaborate on our manuscript. We have clarified the rationale for adopting unsupervised ML clustering in the Methods section (page 10, lines 2–6), emphasizing that this approach was chosen because the study was exploratory and aimed to identify previously unrecognized phenotypes without pre-imposed labels. As the reviewer points out, supervised learning is indeed beneficial for predicting future outcomes. We have already implemented a supervised predictive model in this study, and we have now revised the Methods section to make this clearer and to highlight its role in predicting abdominal pain (AP) occurrence. (page 11, line11–16) 2. Illustrations should be added using cluster analysis, hierarchical structure of the identified clusters, etc. Emerging topics should be visual, so that your audience receives your point of view. Response: We sincerely appreciate this constructive suggestion. As recommended, we have added Condense Tree to illustrate the hierarchical structure of the identified phenotypes (Figure 2B) and plots color-coded by paediatric allergic diseases and representative maternal comorbidities (Figure 2C). 3. The cluster were mainly focused on allergic disease which is not a common in abdominal pain. I would explain this in detail, mentioning this finding from beginning to end. Are you able to explain the algorithm or coding. The result of allergic disease associated with abdominal pain is a bit farfetched with their being more common associations. Response: Thank you for raising this important point regarding the our clustering result. We agree that the clustering procedure should be described in detail, and we have now added the parameters of UMAP and HDBSCAN in S2 Table for reproducibility. We acknowledge that the finding of allergic diseases as a major focus is a novel finding in AP and therefore we have provided a more detailed description of our methods as requested and also in the discussion section as described below. While allergic mechanisms may not represent the main pathophysiology of paediatric AP, previous reports have suggested their involvement, including in irritable bowel syndrome (page 21, line 15 to page 22, line 16). Thus, considering allergic mechanisms as a potential contributor to AP is not without precedent. In our study, the inclusion of a unique cohort with a high prevalence of allergic diseases, particularly among children of Pakistani origin, may have contributed to this finding (page 23, line 17 to page 24, line 16). We believe it is possible that allergic mechanisms play a role in a subgroup of paediatric AP, and this possibility warrants further investigation. In this sense, we consider our study valuable as a foundation for future research. 4. The significance of the population being mostly Pakistan and White could be multifactorial (environmental, etc.) Response: We appreciate the reviewer’s insightful comment. We agree that the ethnic composition of the cohort may have influenced the observed results. In particular, the BiB cohort includes a large proportion of participants of Pakistani origin, a population known to have a higher prevalence of consanguinity, which may contribute to unique genetic characteristics. In addition, cultural and dietary habits may differ from those of the other major group, White participants. Furthermore, as noted earlier, the prevalence of allergic diseases is significantly higher among individuals of Pakistani origin compared with other ethnic groups. These factors have been acknowledged and discussed in the Discussion section (page 23, line 15 to page 24, line 16). 5) The phenotypes seemed unmatched with actual societal groups with abdominal pain or maybe novel new finding. Response: We agree that the identified phenotypes may not align with conventional societal groupings of AP. As noted in our response to Comment 3, we consider this result as potentially novel, generating new research questions linking allergy to paediatric AP and further research to understand the underlying mechanisms is now warranted. We have clarified this point more explicitly in the Discussion section (page 21, line 15 to page 22, line 16). 6) If AI is main objective, this is great but if using the association to come to a conclusion more work is needed. If this is a manuscript on allergic disease, AI and maternal comorbidities, great read but to associate with Abdominal Pain is a bit farfetched. Response: As stated in our previous responses, this is an exploratory study using AI-based clustering, and the findings are not intended to establish definitive causal relationships between allergic diseases, other factors, and AP. We agree that drawing definitive conclusions would require studies with different designs and replication in independent cohorts. Nevertheless, we believe that our study has value in generating novel hypotheses and in highlighting research questions that warrant further investigation in future research. We have added further clarification regarding the future value of our research in our conclusion section on page 27, line 5–14. 7) Your audience needs to be taught ad convinced that AI is appropriate for science and will generate conclusive results. Response: We appreciate this important comment. To help readers better understand the appropriateness of AI in this context, we have expanded the Methods section to provide a clearer description of the AI approach used (page 9, line 17 to page 13, line 8). In addition, we explicitly state that this approach is aimed at generating new hypothesis and research questions to facilitate for future research on this important topic (page 27, line 6–10). Comments from Reviewer 2 The study by Kazuya Takahashi et al. explores pediatric abdominal pain (AP) using machine learning (ML) to identify phenotypes and predict risk factors. Analyzing data from 13,790 children in the Born in Bradford cohort, the researchers identified three AP phenotypes: allergic predisposition, maternal comorbidities, and minimal comorbidities. Allergic diseases and maternal health issues significantly increased the frequency of AP, with 17.6% of children having ≥3 allergic diseases and 25.6% of children with ≥3 maternal comorbidities experiencing AP. A supervised ML model achieved moderate predictive performance (AUC 0.67), highlighting ethnicity, pediatric allergic diseases, and maternal comorbidities as key factors. Risk stratification showed AP rates ranging from 18.9% (<40% probability) to 100% (>60% probability). These findings emphasize the role of genetic, environmental, and maternal influences in the development of AP, offering insights for future research and personalized interventions. The study question is relevant to clinical practice, and the authors conducted a thorough literature review. The analysis of data from 13,790 children provides robust statistical power and generalizability. The major findings are clearly presented, and the research objectives are effectively addressed. The results support the conclusions, and the manuscript is well-written. 1) The authors should discuss in greater detail the gap in existing knowledge regarding AP phenotypes and predictive factors to better justify the need for the study. Response: Thank you for providing us with the opportunity to elaborate on our manuscript. We have expanded the Discussion section to provide a more detailed description of the existing knowledge gap regarding the causes of AP, thereby better justifying the need for this study. In addition, as also noted by Reviewer 1, we have provided a more detailed discussion of the observed association between allergy and AP (page 21, line 15 to page 22, line 16). 2) A brief overview of the methods (e.g., ML clustering and predictive modeling) would help readers understand how the study addresses the research gap. Response: As also noted by Reviewer 1, we have expanded the Methods section to include a brief description of the ML clustering and predictive modeling approaches used in this study (page 9, line 17 to page 13, line 8). We believe that these additions make it clearer why AI was employed and how the study addresses the identified research gap. 3. Summarizing the most important results at the beginning of each subsection in the Results section would help readers quickly grasp the main takeaways. Response: Thank you for this helpful suggestion. We have revised the Results section to include a summary of the key findings at the beginning of each subsection, as recommended. We believe this change improves the readability of the section and allows readers to more easily grasp the main takeaways. 4. The Discussion section could be strengthened by incorporating more data from previous studies to contextualize the findings and highlight how this study advances knowledge on pediatric abdominal pain. Response: Thank you for pointing out this issue. We have addressed this point in the Discussion by considering previous reports alongside our own findings and by outlining the hypothesis that emerges from this study (page 21, line 15 to page 24, line 16). 5. The authors should provide deeper insights into the clinical implications of the identified phenotypes, including how they could inform personalized treatment strategies. Response: We have addressed it in the Discussion by describing potential personalized treatment strategies that could be considered in the future for Phenotype 1 (page 22, line 10–16) and Phenotype 2 (page 22, line 18 to page 23, line 3). 6. More specific recommendations for future studies would be beneficial, such as incorporating pathophysiological data, validating findings in diverse populations, and exploring genetic factors. Response: We appreciate the reviewer’s suggestion. We have incorporated this point into the conclusion, noting the importance of clarifying the underlying pathophysiological mechanisms, validating these results in more diverse populations, and exploring potential genetic and environmental contributors (page 27 line 7–10). Attachments Attachment Submitted filename: response letter_submit.docx https://doi.org/10.1371/journal.pone.0336215.r002
23 Oct 2025 Decision Letter - Hany Abo-Haded, Editor An exploratory machine learning study on paediatric abdominal pain phenotyping and prediction PONE-D-25-34689R1 Dear Dr. Takahashi, 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, Hany Mahmoud Abo-Haded, MD Academic Editor PLOS ONE Additional Editor Comments (optional): Reviewers' comments: Reviewer's Responses to Questions Comments to the Author 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??> Reviewer #1: Yes Reviewer #2: Yes ****** 3. Has the statistical analysis been performed appropriately and rigorously? -->?> Reviewer #1: N/A Reviewer #2: Yes ****** 4. Have the authors made all data underlying the findings in their manuscript fully available??> The PLOS Data policy Reviewer #1: Yes Reviewer #2: Yes ****** 5. Is the manuscript presented in an intelligible fashion and written in standard English??> Reviewer #1: Yes Reviewer #2: Yes ****** Reviewer #1: I appreciate your response to my review, and it reads much easier now for a focused academic reader and also a lay person not in the field. Thanks you for all your work with helping to bring AI to Healthcare. All the best. Reviewer #2: The authors addressed my comments and incorporated my suggestions. In my opinion, the manuscript is ready for publication. ****** 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: Nicole Y Fatheree Reviewer #2: Yes: Rajmohan Dharmaraj, MD ******** https://doi.org/10.1371/journal.pone.0336215.r003
Formally Accepted
Acceptance Letter - Hany Abo-Haded, Editor PONE-D-25-34689R1 PLOS ONE Dear Dr. Takahashi, 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 Professor Hany Mahmoud Abo-Haded Academic Editor PLOS ONE https://doi.org/10.1371/journal.pone.0336215.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 .