HERCULES: an integrative deep-learning framework for predicting RNA-binding propensity and mutation effects at single-residue resolution

HERCULES: an integrative deep-learning framework for predicting RNA-binding propensity and mutation effects at single-residue resolution | 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 Article HERCULES: an integrative deep-learning framework for predicting RNA-binding propensity and mutation effects at single-residue resolution Gian Gaetano Tartaglia, Jonathan Fiorentino, Michele Monti, Alexandros Armaos, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9238243/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract RNA-binding proteins (RBPs) regulate essential aspects of RNA metabolism, yet accurately identifying RNA-binding domains (RBDs) and quantifying the impact of sequence variation on RNA-binding ability remain challenging. Here, we present HERCULES (Hybrid framEwoRk for RNA-binding domain loCalization and mUtation anaLysis using physicochemical and languagE modelS), a unified sequence-based framework for simultaneous RBD localization, global RNA-binding propensity prediction and mutation effect assessment. HERCULES integrates a fine-tuned protein language model with an explicit residue-level physicochemical module, combining global contextual representations with local mutation-sensitive descriptors. On an independent test set, the HERCULES global score discriminates RBPs from non-RBPs with an AUROC of 0.86. At residue resolution, HERCULES outperforms state-of-the-art sequence-based predictors in identifying canonical, non-canonical and putative RBDs across Pfam-annotated proteins. Using a curated dataset of experimentally validated RNA-binding–disrupting mutations, HERCULES correctly classifies 87% of deleterious variants, including single–amino acid substitutions. Evaluation on experimentally resolved protein–RNA complexes further demonstrates robust residue-level performance and improved generalization when contact annotations are augmented with AlphaFold3-predicted complexes. By unifying domain localization and mutation sensitivity within a single sequence-only framework, HERCULES provides a mechanistically interpretable approach for studying RNA–protein interactions. HERCULES is freely available at https://tools.tartaglialab.com/hercules and as an open-source Python package at https://github.com/tartaglialabIIT/hercules.git . Biological sciences/Computational biology and bioinformatics Biological sciences/Biochemistry Biological sciences/Biophysics/Computational biophysics Full Text Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryTableS1.xlsx Supplementary Table 1. SupplementaryTableS2.xlsx Supplementary Table 2. SupplementaryTableS3.xlsx Supplementary Table 3. SupplementaryTableS4.xlsx Supplementary Table 4. SupplementaryTableS5.xlsx Supplementary Table 5. HERCULESNATMethSupplementary.docx Supplementary Materials Cite Share Download PDF Status: Under Review Version 1 posted 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-9238243","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":623203476,"identity":"106dc6bb-51c8-4d0a-aed2-7cdfcd4ce395","order_by":0,"name":"Gian Gaetano Tartaglia","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA70lEQVRIiWNgGAWjYDACZgST8UACgw0PG5B1gLCWBAYGHpDKBIY0HjagngP49SBrYWA4zMBAyBp5d+ZnDxh/2OXbsx+/cOBBzXkZPvkGxsMf8GgxPMxmbsCQkGzZw5NTcCDh2G3CDjNsZjCTYEhgNuBhyEk4kMBGlBb2b0At9QY8/G+AWv6dI6xFnpkHZMthAx6J9AMHEtsOENZiwMxTJpGQdtyA58YbhgOJfclALYkNB87gs6X/+DaJDzbVBuz96Q8f/vhmZy/ffPjwhwp8toCckABm8hhAxRgb8GgA2oKQZn+AV+UoGAWjYBSMXAAASmdK89OC/W0AAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0001-7524-6310","institution":"Istituto Italiano di Tecnologia","correspondingAuthor":true,"prefix":"","firstName":"Gian","middleName":"Gaetano","lastName":"Tartaglia","suffix":""},{"id":623203477,"identity":"f0411e82-74dd-40a1-a328-bf176456d1ce","order_by":1,"name":"Jonathan Fiorentino","email":"","orcid":"","institution":"Istituto Italiano di Tecnologia","correspondingAuthor":false,"prefix":"","firstName":"Jonathan","middleName":"","lastName":"Fiorentino","suffix":""},{"id":623203478,"identity":"ceca3e2e-a056-48f6-a862-c4ee5d625edc","order_by":2,"name":"Michele Monti","email":"","orcid":"","institution":"Istituto Italiano di Tecnologia","correspondingAuthor":false,"prefix":"","firstName":"Michele","middleName":"","lastName":"Monti","suffix":""},{"id":623203479,"identity":"e671c2ef-4c24-476c-a984-d9333ebf6df8","order_by":3,"name":"Alexandros Armaos","email":"","orcid":"","institution":"Italian Institute of Technology","correspondingAuthor":false,"prefix":"","firstName":"Alexandros","middleName":"","lastName":"Armaos","suffix":""},{"id":623203480,"identity":"9e36650f-15a4-48d5-8a9a-d0465db7c054","order_by":4,"name":"Dimitrios Miltiadis-Vrachnos","email":"","orcid":"","institution":"University of Rome Sapienza","correspondingAuthor":false,"prefix":"","firstName":"Dimitrios","middleName":"","lastName":"Miltiadis-Vrachnos","suffix":""},{"id":623203481,"identity":"9b8a648d-a9ff-4327-aa8a-368f6363ca01","order_by":5,"name":"Lorenzo Di Rienzo","email":"","orcid":"","institution":"University of Rome Sapienza","correspondingAuthor":false,"prefix":"","firstName":"Lorenzo","middleName":"Di","lastName":"Rienzo","suffix":""}],"badges":[],"createdAt":"2026-03-26 22:20:43","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9238243/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9238243/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108170920,"identity":"ea80a099-2b0a-430d-92af-9fa02c4f9459","added_by":"auto","created_at":"2026-04-30 07:00:46","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":736594,"visible":true,"origin":"","legend":"Supplementary Table 1.","description":"","filename":"SupplementaryTableS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9238243/v1/8d98f5862c60c0f282c119cf.xlsx"},{"id":108182869,"identity":"1d5a1228-4264-4355-a3b6-cd20b463cc6b","added_by":"auto","created_at":"2026-04-30 08:59:39","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":823460,"visible":true,"origin":"","legend":"Supplementary Table 2.","description":"","filename":"SupplementaryTableS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9238243/v1/b57a1552e938871d020c31a0.xlsx"},{"id":108170922,"identity":"d7375a9f-a9e1-4dfd-9b5d-adb94cabe81a","added_by":"auto","created_at":"2026-04-30 07:00:46","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":31385,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Table 3.\u003c/p\u003e","description":"","filename":"SupplementaryTableS3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9238243/v1/9f94b94aef2575d480477aa5.xlsx"},{"id":108182839,"identity":"8545d1bc-bd0f-410e-8cd6-ba3d470dde07","added_by":"auto","created_at":"2026-04-30 08:59:37","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":22828,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Table 4.\u003c/p\u003e","description":"","filename":"SupplementaryTableS4.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9238243/v1/ff4496b92a134ea87f56ea82.xlsx"},{"id":108170925,"identity":"cedf26fb-9876-49ca-a6fb-04cb2a94d3cf","added_by":"auto","created_at":"2026-04-30 07:00:47","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":43340,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Table 5.\u003c/p\u003e","description":"","filename":"SupplementaryTableS5.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9238243/v1/f101506aff719051a27659ea.xlsx"},{"id":108182715,"identity":"b85371ce-8b62-4e3f-a374-2b3cbaf45d0c","added_by":"auto","created_at":"2026-04-30 08:59:31","extension":"docx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":4691103,"visible":true,"origin":"","legend":"Supplementary Materials","description":"","filename":"HERCULESNATMethSupplementary.docx","url":"https://assets-eu.researchsquare.com/files/rs-9238243/v1/6458cedfc906393cb30cd602.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"HERCULES: an integrative deep-learning framework for predicting RNA-binding propensity and mutation effects at single-residue resolution","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-9238243/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9238243/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"RNA-binding proteins (RBPs) regulate essential aspects of RNA metabolism, yet accurately identifying RNA-binding domains (RBDs) and quantifying the impact of sequence variation on RNA-binding ability remain challenging. Here, we present HERCULES (Hybrid framEwoRk for RNA-binding domain loCalization and mUtation anaLysis using physicochemical and languagE modelS), a unified sequence-based framework for simultaneous RBD localization, global RNA-binding propensity prediction and mutation effect assessment. HERCULES integrates a fine-tuned protein language model with an explicit residue-level physicochemical module, combining global contextual representations with local mutation-sensitive descriptors. On an independent test set, the HERCULES global score discriminates RBPs from non-RBPs with an AUROC of 0.86. At residue resolution, HERCULES outperforms state-of-the-art sequence-based predictors in identifying canonical, non-canonical and putative RBDs across Pfam-annotated proteins. Using a curated dataset of experimentally validated RNA-binding–disrupting mutations, HERCULES correctly classifies 87% of deleterious variants, including single–amino acid substitutions. Evaluation on experimentally resolved protein–RNA complexes further demonstrates robust residue-level performance and improved generalization when contact annotations are augmented with AlphaFold3-predicted complexes. By unifying domain localization and mutation sensitivity within a single sequence-only framework, HERCULES provides a mechanistically interpretable approach for studying RNA–protein interactions. HERCULES is freely available at https://tools.tartaglialab.com/hercules and as an open-source Python package at https://github.com/tartaglialabIIT/hercules.git.","manuscriptTitle":"HERCULES: an integrative deep-learning framework for predicting RNA-binding propensity and mutation effects at single-residue resolution","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-30 07:00:41","doi":"10.21203/rs.3.rs-9238243/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"97481f3e-5700-4791-8e18-8f712ed4f19e","owner":[],"postedDate":"April 30th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":66323464,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":66323465,"name":"Biological sciences/Biochemistry"},{"id":66323466,"name":"Biological sciences/Biophysics/Computational biophysics"}],"tags":[],"updatedAt":"2026-04-30T07:00:42+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-30 07:00:41","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9238243","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9238243","identity":"rs-9238243","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}