SOT-RM: An equivariant shortcut and optimal transport-based flow model for exploring chemical reaction mechanisms | 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 SOT-RM: An equivariant shortcut and optimal transport-based flow model for exploring chemical reaction mechanisms Mingyuan Xu, Bowen Li, Zhaojia Dong, Yuxinxin Chen, Pavlo Dral, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8836310/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 Accurately modeling three-dimensional reaction mechanisms remains a critical bottleneck in chemistry, as reliable transition states and minimum energy pathways require expensive quantum chemistry calculations with expert-intensive setup. Here we introduce SOT-RM, a unified SE (3)-equivariant generative framework that integrates the direct prediction of transition state (TS) geometries and minimum energy reaction pathways. SOT-RM introduces few-step shortcut-flow with optimal transport-based noise alignment, overcoming the integration error and noise-geometry mismatch that fundamentally limit conventional flow-based models. On Transition1x, SOT-RM achieves 0.076 Å mean TS RMSD and 1.45 kcal/mol mean ΔE relative to DFT (lowest-energy selection from 4 samples), with 79.5% of predicted transition states exhibiting the correct single imaginary frequency. Beyond geometries and energies, SOT-RM captures electronic redistribution along reaction coordinates (Wiberg bond-order RMSD 0.039). Inference requires ~0.06 s per reaction, ~10³× faster than DFT, with strong generalization demonstrated on the external Transition1+ benchmark. By enabling automated, large-scale mechanistic exploration, SOT-RM bridges static quantum chemistry and dynamic reaction processes. Physical sciences/Chemistry/Cheminformatics Physical sciences/Chemistry/Theoretical chemistry/Reaction mechanisms Physical sciences/Chemistry/Organic chemistry/Reaction mechanisms Full Text Additional Declarations There is NO Competing Interest. Supplementary Files Appendix.docx 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-8836310","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":604813137,"identity":"2ba3be93-17c5-4310-a67b-31e23d9c7b5e","order_by":0,"name":"Mingyuan Xu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+UlEQVRIiWNgGAWjYDACZlSuhByIPPCAFC3GYC0JpNia2AAi8WnRbec9/Jq3zS6PQSLHgPFHjUX6/LDDD4G22MnpNmDXYnaYL82aty25mIHnjAEzzzGJ3I230wyAWpKNzQ7g0sJjZszbxpzYwN5jwMzYANQyOwGk5UDiNvxa6hMbmHkMGH82SKQbzk7/QEiL8WPetsNgWxh4GyQS5KVzCNvCOOfc8cQ2nmMFh4F+MdwgnVNwIMEAj1/OnzH+8KasOrFfInnjwx81dfLys9M3f/hQYSeHSwsQsEnxgEggBqsxgJA4lYMA88cfyFz5BryqR8EoGAWjYAQCAC55WlFwBOnVAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0003-0463-4249","institution":"Guangzhou National Laboratory","correspondingAuthor":true,"prefix":"","firstName":"Mingyuan","middleName":"","lastName":"Xu","suffix":""},{"id":604813138,"identity":"4d434198-ec50-4994-9d52-c219d3f3eff2","order_by":1,"name":"Bowen Li","email":"","orcid":"","institution":"Shanghai Engineering Research Center of Molecular Therapeutics \u0026 New Drug Development, School of Chemistry and Molecular Engineering, East China Normal University","correspondingAuthor":false,"prefix":"","firstName":"Bowen","middleName":"","lastName":"Li","suffix":""},{"id":604813139,"identity":"d1427f22-c20e-4c9e-964e-0d2e145a1ddb","order_by":2,"name":"Zhaojia Dong","email":"","orcid":"","institution":"Institute of Atomic and Molecular Physics, Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Zhaojia","middleName":"","lastName":"Dong","suffix":""},{"id":604813140,"identity":"1e920379-560c-4eef-b727-889083f30d17","order_by":3,"name":"Yuxinxin Chen","email":"","orcid":"https://orcid.org/0000-0001-8398-5690","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Yuxinxin","middleName":"","lastName":"Chen","suffix":""},{"id":604813141,"identity":"c544fa04-703a-4860-b095-1a13e3a0f2e9","order_by":4,"name":"Pavlo Dral","email":"","orcid":"https://orcid.org/0000-0002-2975-9876","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Pavlo","middleName":"","lastName":"Dral","suffix":""},{"id":604813142,"identity":"c0a6b7df-dfc1-454f-902b-6a93bb9feebf","order_by":5,"name":"Tong Zhu","email":"","orcid":"https://orcid.org/0000-0001-7472-3736","institution":"East China Normal University","correspondingAuthor":false,"prefix":"","firstName":"Tong","middleName":"","lastName":"Zhu","suffix":""},{"id":604813143,"identity":"5ff7ad87-bec0-4f17-b20e-f23fc7bf8a43","order_by":6,"name":"Hongming Chen","email":"","orcid":"https://orcid.org/0000-0002-8065-8333","institution":"Guangzhou National Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Hongming","middleName":"","lastName":"Chen","suffix":""}],"badges":[],"createdAt":"2026-02-10 04:41:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8836310/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8836310/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107707844,"identity":"09cbdbaf-a3cc-4383-b9a9-1cc313489f83","added_by":"auto","created_at":"2026-04-24 09:21:15","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1939677,"visible":true,"origin":"","legend":"","description":"","filename":"SCMechnismEngsubmission.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8836310/v1_covered_7e0543fb-01f0-46f0-95de-4dd84c48cca0.pdf"},{"id":107683348,"identity":"95dfebc6-35fd-4322-9ec9-329c32da5663","added_by":"auto","created_at":"2026-04-24 03:28:06","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":45299,"visible":true,"origin":"","legend":"","description":"","filename":"Appendix.docx","url":"https://assets-eu.researchsquare.com/files/rs-8836310/v1/a2bbc0cc4fe06403775c3e32.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"SOT-RM: An equivariant shortcut and optimal transport-based flow model for exploring chemical reaction mechanisms","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"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":"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-8836310/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8836310/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Accurately modeling three-dimensional reaction mechanisms remains a critical bottleneck in chemistry, as reliable transition states and minimum energy pathways require expensive quantum chemistry calculations with expert-intensive setup. Here we introduce SOT-RM, a unified SE (3)-equivariant generative framework that integrates the direct prediction of transition state (TS) geometries and minimum energy reaction pathways. SOT-RM introduces few-step shortcut-flow with optimal transport-based noise alignment, overcoming the integration error and noise-geometry mismatch that fundamentally limit conventional flow-based models. On Transition1x, SOT-RM achieves 0.076 Å mean TS RMSD and 1.45 kcal/mol mean ΔE relative to DFT (lowest-energy selection from 4 samples), with 79.5% of predicted transition states exhibiting the correct single imaginary frequency. Beyond geometries and energies, SOT-RM captures electronic redistribution along reaction coordinates (Wiberg bond-order RMSD 0.039). Inference requires ~0.06 s per reaction, ~10³× faster than DFT, with strong generalization demonstrated on the external Transition1+ benchmark. By enabling automated, large-scale mechanistic exploration, SOT-RM bridges static quantum chemistry and dynamic reaction processes.","manuscriptTitle":"SOT-RM: An equivariant shortcut and optimal transport-based flow model for exploring chemical reaction mechanisms","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-24 03:28:02","doi":"10.21203/rs.3.rs-8836310/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":"abba41cc-46ea-4287-a9d0-48444d9fc33f","owner":[],"postedDate":"April 24th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":64364272,"name":"Physical sciences/Chemistry/Cheminformatics"},{"id":64364273,"name":"Physical sciences/Chemistry/Theoretical chemistry/Reaction mechanisms"},{"id":64364274,"name":"Physical sciences/Chemistry/Organic chemistry/Reaction mechanisms"}],"tags":[],"updatedAt":"2026-04-24T03:28:02+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-24 03:28:02","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8836310","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8836310","identity":"rs-8836310","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}
Text is read by the "Ask this paper" AI Q&A widget below.
Extraction quality varies by source — PMC NXML preserves structure
cleanly, OA-HTML may include some navigation residue, and OA-PDF can
have broken hyphenation. The publisher copy
(via DOI)
is the canonical version.