Framework for evaluating explainable AI in antimicrobial drug discovery

Framework for evaluating explainable AI in antimicrobial drug discovery | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Method Article Framework for evaluating explainable AI in antimicrobial drug discovery Abdulmujeeb T. Onawole, Mark A. T. Blaskovich, Johannes Zuegg This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8615784/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract Explainable artificial intelligence (XAI) methods for molecular property prediction lack standardized evaluation criteria, preventing widespread deployment in drug development and hit optimisation, where proper understanding of structure-activity relationship is essential. We developed an evaluation framework for XAI using fragment-based explainability tests to compare XAI with different molecular representation and challenge the different XAI approaches for proper explanation of activity cliffs. The evaluation methods include essential scaffold recognition, scaffold sensitivity and substructure specificity for explaining activity cliff, and technical evaluation on model robustness and consistency. Using a curated dataset of antibiotic molecules we established three XAI models with fundamentally different molecular representation: Random Forest on chemical features using SHAP, CNN on sequence-based SMILES using token occlusion, and RGCN on molecular graphs with substructure masking. Together with detailed case study, we evaluated the explainability behaviours and quality of the different XAI approaches and highlighted their limitations. While all XAI approaches displayed good predictive and scaffold recognition capabilities, and comparable robustness and consistency, they displayed quite different explainability behaviour for activity cliffs, revealing their different utility for medicinal chemistry. Machine learning Deep Neural Networks Explainable AI Antibacterial prediction Drug development Structure-activity relationship Chemoinformatic Molecular decomposition Full Text Additional Declarations No competing interests reported. Supplementary Files XAIFrameworkJZueggSupportingInformation.docx Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 20 Feb, 2026 Reviews received at journal 16 Feb, 2026 Reviews received at journal 12 Feb, 2026 Reviewers agreed at journal 01 Feb, 2026 Reviewers agreed at journal 01 Feb, 2026 Reviewers invited by journal 26 Jan, 2026 Editor assigned by journal 20 Jan, 2026 Submission checks completed at journal 20 Jan, 2026 First submitted to journal 16 Jan, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8615784","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Method Article","associatedPublications":[],"authors":[{"id":581971663,"identity":"89c6dc5b-a97b-4a44-86c8-0dc2655c6a86","order_by":0,"name":"Abdulmujeeb T. Onawole","email":"","orcid":"","institution":"University of Queensland","correspondingAuthor":false,"prefix":"","firstName":"Abdulmujeeb","middleName":"T.","lastName":"Onawole","suffix":""},{"id":581971664,"identity":"73106a9c-eee8-490a-a77b-f6e46c4042e2","order_by":1,"name":"Mark A. T. Blaskovich","email":"","orcid":"","institution":"University of Queensland","correspondingAuthor":false,"prefix":"","firstName":"Mark","middleName":"A. T.","lastName":"Blaskovich","suffix":""},{"id":581971665,"identity":"fc63f6b9-4ded-4f77-b75d-a3705848ef56","order_by":2,"name":"Johannes Zuegg","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzElEQVRIiWNgGAWjYBACxgYILQcmeUjRYky8FhhIbCBaC3MD+zOJD79q0zccT2B88LaNQd7gAGGHpUnO7Dueu+HMA2bDuW0MhhuI0HJMmrfnWO6GGwls0rxtDIxEaGFsk/7bcyzd4EYC+2+gFnsitDCzSTP8qEkAamFjBmpJJKylmY3ZsrfhgOHMMw+bJeeck0ieSUiLYXv7wxs//tTJ8x1PPvjhTZmNbR9BLc0gq9oOM0CjRoKAeiCQB5N/6oBEAmHVo2AUjIJRMDIBAEp0RBVYeUagAAAAAElFTkSuQmCC","orcid":"","institution":"University of Queensland","correspondingAuthor":true,"prefix":"","firstName":"Johannes","middleName":"","lastName":"Zuegg","suffix":""}],"badges":[],"createdAt":"2026-01-16 08:21:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8615784/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8615784/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":101398651,"identity":"5b9d8d77-59f3-4897-b3f0-f0d485eb8827","added_by":"auto","created_at":"2026-01-29 09:43:30","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1331248,"visible":true,"origin":"","legend":"","description":"","filename":"XAIFrameworkJZuegg19Jan2026.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8615784/v1_covered_ed7496aa-af78-471d-a656-35c5d42bf5fb.pdf"},{"id":101382300,"identity":"9217d2d7-5d6f-49d6-8ea4-b7f5ed840afe","added_by":"auto","created_at":"2026-01-29 06:26:47","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":6754852,"visible":true,"origin":"","legend":"","description":"","filename":"XAIFrameworkJZueggSupportingInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-8615784/v1/0d16fe9679cb9b1bc1220afb.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Framework for evaluating explainable AI in antimicrobial drug discovery","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"journal-of-cheminformatics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"chin","sideBox":"Learn more about [Journal of Cheminformatics](https://jcheminf.biomedcentral.com/)","snPcode":"13321","submissionUrl":"https://submission.nature.com/new-submission/13321/3","title":"Journal of Cheminformatics","twitterHandle":"@jcheminf","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Machine learning, Deep Neural Networks, Explainable AI, Antibacterial prediction, Drug development, Structure-activity relationship, Chemoinformatic, Molecular decomposition","lastPublishedDoi":"10.21203/rs.3.rs-8615784/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8615784/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eExplainable artificial intelligence (XAI) methods for molecular property prediction lack standardized evaluation criteria, preventing widespread deployment in drug development and hit optimisation, where proper understanding of structure-activity relationship is essential. We developed an evaluation framework for XAI using fragment-based explainability tests to compare XAI with different molecular representation and challenge the different XAI approaches for proper explanation of activity cliffs. The evaluation methods include essential scaffold recognition, scaffold sensitivity and substructure specificity for explaining activity cliff, and technical evaluation on model robustness and consistency. Using a curated dataset of antibiotic molecules we established three XAI models with fundamentally different molecular representation: Random Forest on chemical features using SHAP, CNN on sequence-based SMILES using token occlusion, and RGCN on molecular graphs with substructure masking. Together with detailed case study, we evaluated the explainability behaviours and quality of the different XAI approaches and highlighted their limitations. While all XAI approaches displayed good predictive and scaffold recognition capabilities, and comparable robustness and consistency, they displayed quite different explainability behaviour for activity cliffs, revealing their different utility for medicinal chemistry.\u003c/p\u003e","manuscriptTitle":"Framework for evaluating explainable AI in antimicrobial drug discovery","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-29 06:26:17","doi":"10.21203/rs.3.rs-8615784/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-02-20T12:23:04+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-16T17:39:34+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-12T06:21:09+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"283035481311430665196780402599444499613","date":"2026-02-02T04:11:13+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"103665064111117298942797720977214626646","date":"2026-02-01T10:43:46+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-01-26T15:15:48+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-21T01:45:27+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-01-20T06:01:42+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of Cheminformatics","date":"2026-01-16T05:55:45+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"journal-of-cheminformatics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"chin","sideBox":"Learn more about [Journal of Cheminformatics](https://jcheminf.biomedcentral.com/)","snPcode":"13321","submissionUrl":"https://submission.nature.com/new-submission/13321/3","title":"Journal of Cheminformatics","twitterHandle":"@jcheminf","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"808f84ff-2390-400a-9b13-0f113214181a","owner":[],"postedDate":"January 29th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-07T13:11:22+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-29 06:26:17","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8615784","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8615784","identity":"rs-8615784","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}