Reaction Center and Class Prediction via Cross-Attentive Multi-Task Learning

A cross-attentive multi-task graph learning framework for chemical reaction modeling | 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 Research Article A cross-attentive multi-task graph learning framework for chemical reaction modeling Maryam Astero, Anchen Li, Elena Casiraghi, Juho Rousu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6977897/v2 This work is licensed under a CC BY 4.0 License Status: Posted Version 2 posted You are reading this latest preprint version Show more versions Abstract Understanding chemical reactions requires bridging fine-grained molecular edits with broader semantic context. Reaction mechanisms are determined not only by local atom–bond transformations but also by the global reaction class. However, most existing approaches treat these tasks separately or rely on external atom-mapping tools, introducing noise and limiting end-to-end learnability. We introduce MARCC (Mapping-Assisted Reaction Center and Classification), a multi-task graph neural network that jointly predicts atom mappings, reaction centers, and reaction classes within a unified architecture. MARCC integrates three key innovations: (i) a mapping-guided cross-attention mechanism that aligns reactants and products for local edit detection, (ii) a dual-graph design that explicitly reasons about bond-level transformations, and (iii) pooled product embeddings for global reaction classification. On the USPTO-50K benchmark, MARCC achieves state-of-the-art results when trained with both reactants and products, including 98.2% atom mapping accuracy, 99.1% Top-1 edit localization accuracy, and 97.2% reaction classification accuracy. Even under the products-only setting, MARCC delivers competitive performance comparable to specialized baselines. Ablation studies confirm the value of mapping-guided attention and multi-task supervision, which enhance both predictive accuracy and interpretability. By unifying atom-level alignment, local reactivity, and global classification, MARCC provides a structured and interpretable framework for reaction understanding. Beyond benchmarks, MARCC has the potential to support applications in reaction annotation, template discovery, and mechanism inference; with additional domain-specific modeling and data, it could be extended to biochemical domains such as enzyme-catalyzed transformations and metabolic pathway modeling. graph neural networks multi-task learning atom mapping reaction center identification reaction classification cross-graph attention chemical reaction prediction Full Text Additional Declarations The authors declare no competing interests. Supplementary Files supplementary.pdf Cite Share Download PDF Status: Posted Version 2 posted You are reading this latest preprint version Show more versions 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-6977897","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":479354012,"identity":"58a6dec9-98b1-4c69-bee6-bd5faaeb384b","order_by":0,"name":"Maryam Astero","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA9ElEQVRIiWNgGAWjYPCDCgYGPjBtQIxqNhBxBkiDGGeI1sLYBtOCR6Fu+9lnEj9zGOTl5zcf+/Bxnp08m3zzAYYDBbi1mJ1JN5Ps3cZguOEYW/LMmduSDdvY2BIYDuBxmNmBNDYJ3m0MjBvYeIyZebcxM7ax8Rgwf8Cn5fwzNsm/2xjs57fxf2bmnVNvD9KC35YbaWzSQFsSG47xMDPzNhxOJELLM2Zr2W0SyRuOpRkzzjh2PLmNLS3hAF4t59MYb77dZmM7v/nwY4YPNdW2/cyHDz448Ae3FiBgkWBgkEAVOoBXAwMD8wcCCkbBKBgFo2CkAwA6fkr870vMOgAAAABJRU5ErkJggg==","orcid":"","institution":"Aalto University","correspondingAuthor":true,"prefix":"","firstName":"Maryam","middleName":"","lastName":"Astero","suffix":""},{"id":479354013,"identity":"e4a71a28-81f5-4e15-993a-429425e831ed","order_by":1,"name":"Anchen Li","email":"","orcid":"","institution":"Aalto University","correspondingAuthor":false,"prefix":"","firstName":"Anchen","middleName":"","lastName":"Li","suffix":""},{"id":479354014,"identity":"8ba69d0d-0d48-43b8-9248-4f9a535d1160","order_by":2,"name":"Elena Casiraghi","email":"","orcid":"","institution":"University of Milan","correspondingAuthor":false,"prefix":"","firstName":"Elena","middleName":"","lastName":"Casiraghi","suffix":""},{"id":479354015,"identity":"16355283-7d7e-4db2-8f28-4115b69ca0f0","order_by":3,"name":"Juho Rousu","email":"","orcid":"","institution":"Aalto University","correspondingAuthor":false,"prefix":"","firstName":"Juho","middleName":"","lastName":"Rousu","suffix":""}],"badges":[],"createdAt":"2025-06-25 21:08:13","currentVersionCode":2,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-6977897/v2","doiUrl":"https://doi.org/10.21203/rs.3.rs-6977897/v2","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":90192790,"identity":"dbafa582-a583-4223-82bc-0896368e3648","added_by":"auto","created_at":"2025-08-29 16:13:49","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1108941,"visible":true,"origin":"","legend":"","description":"","filename":"Aunifiedgraphbasedframeworkforchemicalreactionpredictionclean.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6977897/v2_covered_e245b0b7-bde5-416b-9821-f5c7536fe85d.pdf"},{"id":90191881,"identity":"3a791119-d8a6-4ed8-bc17-c59774ce978a","added_by":"auto","created_at":"2025-08-29 15:57:46","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":380462,"visible":true,"origin":"","legend":"","description":"","filename":"supplementary.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6977897/v2/d7e25183f98be22f9b5eac61.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eA cross-attentive multi-task graph learning framework for chemical reaction modeling\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"graph neural networks, multi-task learning, atom mapping, reaction center identification, reaction classification, cross-graph attention, chemical reaction prediction","lastPublishedDoi":"10.21203/rs.3.rs-6977897/v2","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6977897/v2","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eUnderstanding chemical reactions requires bridging fine-grained molecular edits with broader semantic context. Reaction mechanisms are determined not only by local atom–bond transformations but also by the global reaction class. However, most existing approaches treat these tasks separately or rely on external atom-mapping tools, introducing noise and limiting end-to-end learnability. We introduce MARCC (Mapping-Assisted Reaction Center and Classification), a multi-task graph neural network that jointly predicts atom mappings, reaction centers, and reaction classes within a unified architecture. MARCC integrates three key innovations: (i) a mapping-guided cross-attention mechanism that aligns reactants and products for local edit detection, (ii) a dual-graph design that explicitly reasons about bond-level transformations, and (iii) pooled product embeddings for global reaction classification. On the USPTO-50K benchmark, MARCC achieves state-of-the-art results when trained with both reactants and products, including 98.2% atom mapping accuracy, 99.1% Top-1 edit localization accuracy, and 97.2% reaction classification accuracy. Even under the products-only setting, MARCC delivers competitive performance comparable to specialized baselines. Ablation studies confirm the value of mapping-guided attention and multi-task supervision, which enhance both predictive accuracy and interpretability. By unifying atom-level alignment, local reactivity, and global classification, MARCC provides a structured and interpretable framework for reaction understanding. Beyond benchmarks, MARCC has the potential to support applications in reaction annotation, template discovery, and mechanism inference; with additional domain-specific modeling and data, it could be extended to biochemical domains such as enzyme-catalyzed transformations and metabolic pathway modeling.\u003c/p\u003e","manuscriptTitle":"A cross-attentive multi-task graph learning framework for chemical reaction modeling","msid":"","msnumber":"","nonDraftVersions":[{"code":2,"date":"2025-08-29 15:57:42","doi":"10.21203/rs.3.rs-6977897/v2","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}},{"code":1,"date":"2025-07-02 04:04:03","doi":"10.21203/rs.3.rs-6977897/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"151d892b-71ea-4ee2-8563-7f8895fc1c28","owner":[],"postedDate":"August 29th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-07-31T14:39:05+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-29 15:57:42","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v2","identity":"rs-6977897","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6977897","identity":"rs-6977897","version":["v2"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}