Explicit Dynamic Cross-Strand Interactions for DNA Sequence Language Modeling

Explicit Dynamic Cross-Strand Interactions for DNA Sequence Language 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 Article Explicit Dynamic Cross-Strand Interactions for DNA Sequence Language Modeling Xiao Luo, Cheng Yang, Yuansheng Liu, Lei Ling, Fengxin Li, Changjian Chen, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8332638/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 Large language models are driving a transformation in the functional interpretation of genomic sequences. Methodologically, existing DNA sequence language models can be broadly categorized into two types: one that treats DNA sequence as unidirectional text for single-strand modeling, and another that achieves reverse-complement symmetry through data augmentation or model equivariance for static double-strand modeling. Both approaches predominantly approximate double-strand interactions in an implicit and static manner, struggling to capture context-driven cross-strand information exchange during sequence representation learning. In reality, double-strand information exchange is not an isolated event but is regulated by continuous physical coupling, functional synergy, and information transfer—a mechanism fundamental to genomic function. Based on this, we propose CrossDNA, an explicit and dynamic language model for DNA cross-strand modeling. Specifically, CrossDNA employs a dual-branch architecture with rotating input of double-strand sequence data to simulate the continuous information flow in the DNA double helix, establishes inter-strand communication via a lightweight TokenBridge module, and incorporates Comba with window-sliding attention (SWA) to capture long-range dependencies, while maintaining reverse-complement equivalence and stabilizing single-strand contextual semantics through self-distillation and consistency constraints from a branch teacher model. On tasks such as classification, regression, and representation, CrossDNA achieves consistent performance improvements and significantly enhances model robustness to sequence orientation, particularly in enhancer prediction where it more readily identifies features with clear biological significance. On multiple benchmarks we evaluated, CrossDNA, with a model size of only a few million parameters, matches or surpasses the performance of large models with hundreds of millions of parameters, substantially reducing training and inference costs and demonstrating high parameter efficiency and usability. Overall, CrossDNA advances DNA representation from implicit, static approximation to explicit, dynamic systematic modeling, signaling the orientation for a new generation of DNA language models and laying the foundation for deeper analysis of genomic structure and function. Biological sciences/Computational biology and bioinformatics/Computational models Biological sciences/Biological techniques/Bioinformatics Full Text Additional Declarations There is NO Competing Interest. Supplementary Files Supplementary.pdf Supplementary information 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-8332638","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":570278454,"identity":"dac4c96e-3685-4f02-9757-ed2494e424e6","order_by":0,"name":"Xiao Luo","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAq0lEQVRIiWNgGAWjYDACdiD+2ABmGhCphZmBgXEmyVqYeUnSYnCYx/Cx7Y47iQ3szdskGGruENYi2cxjbJx75lliA8+xMgmGY88Ia+Fn5jGTzm07nNggkWMmwdhwmLAWNpAWS5AW+TdEagHbwgi2hYdILZLNbMWGvW2Hjdt40ootEo4RocXgePPGBz/bDsv2sx/eeONDDRFa4IANRCSQoGEUjIJRMApGAR4AAOhfMtYIoG4XAAAAAElFTkSuQmCC","orcid":"","institution":"Hunan University","correspondingAuthor":true,"prefix":"","firstName":"Xiao","middleName":"","lastName":"Luo","suffix":""},{"id":570278455,"identity":"5f9d5b08-89d9-4d61-9dbb-6bdcd3c63f47","order_by":1,"name":"Cheng Yang","email":"","orcid":"","institution":"Hunan University","correspondingAuthor":false,"prefix":"","firstName":"Cheng","middleName":"","lastName":"Yang","suffix":""},{"id":570278456,"identity":"934be421-c881-48b3-ae12-c9647ca83034","order_by":2,"name":"Yuansheng Liu","email":"","orcid":"https://orcid.org/0000-0002-7680-3155","institution":"College of Computer Science and Electronic Engineering, Hunan University","correspondingAuthor":false,"prefix":"","firstName":"Yuansheng","middleName":"","lastName":"Liu","suffix":""},{"id":570278457,"identity":"b750f5b3-8793-4f44-b6cc-a7ad2495544a","order_by":3,"name":"Lei Ling","email":"","orcid":"https://orcid.org/0000-0001-6426-533X","institution":"Hunan University","correspondingAuthor":false,"prefix":"","firstName":"Lei","middleName":"","lastName":"Ling","suffix":""},{"id":570278458,"identity":"9b8f2595-90f1-4e3e-9c5e-c42934f91f68","order_by":4,"name":"Fengxin Li","email":"","orcid":"","institution":"Hunan University","correspondingAuthor":false,"prefix":"","firstName":"Fengxin","middleName":"","lastName":"Li","suffix":""},{"id":570278459,"identity":"a00af6c4-3723-4758-a062-d157c6ba7bb6","order_by":5,"name":"Changjian Chen","email":"","orcid":"","institution":"Hunan University","correspondingAuthor":false,"prefix":"","firstName":"Changjian","middleName":"","lastName":"Chen","suffix":""},{"id":570278460,"identity":"20e7df90-053a-4f62-ae9a-a8a7ecf25d45","order_by":6,"name":"Long Wang","email":"","orcid":"","institution":"Hunan University","correspondingAuthor":false,"prefix":"","firstName":"Long","middleName":"","lastName":"Wang","suffix":""},{"id":570278461,"identity":"dd42e6cd-5d6c-420b-ad72-0aca9f39c403","order_by":7,"name":"Feng Yu","email":"","orcid":"","institution":"Hunan University","correspondingAuthor":false,"prefix":"","firstName":"Feng","middleName":"","lastName":"Yu","suffix":""},{"id":570278462,"identity":"e37b77ad-b6f1-45a4-a111-69f68dee77db","order_by":8,"name":"Liang Qiao","email":"","orcid":"","institution":"Taichu (Wuxi) Electronic Technology Co., Ltd","correspondingAuthor":false,"prefix":"","firstName":"Liang","middleName":"","lastName":"Qiao","suffix":""},{"id":570278463,"identity":"88eb6011-6fa3-4b19-b65a-6a892efb50a8","order_by":9,"name":"Xiangxiang Zeng","email":"","orcid":"https://orcid.org/0000-0003-1081-7658","institution":"Hunan University","correspondingAuthor":false,"prefix":"","firstName":"Xiangxiang","middleName":"","lastName":"Zeng","suffix":""},{"id":570278464,"identity":"3303d3d0-d517-4cf8-ace7-8e7ca34e5f0e","order_by":10,"name":"Kenli Li","email":"","orcid":"","institution":"Hunan University","correspondingAuthor":false,"prefix":"","firstName":"Kenli","middleName":"","lastName":"Li","suffix":""},{"id":570278465,"identity":"e7ddc53d-0c51-4fd8-8a2b-e5ce006e3da4","order_by":11,"name":"Alexander Schönhuth","email":"","orcid":"https://orcid.org/0000-0003-3529-0856","institution":"Bielefeld University","correspondingAuthor":false,"prefix":"","firstName":"Alexander","middleName":"","lastName":"Schönhuth","suffix":""}],"badges":[],"createdAt":"2025-12-11 05:05:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8332638/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8332638/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99798706,"identity":"1bf2a374-95e1-4425-bb75-8552a88b3c2c","added_by":"auto","created_at":"2026-01-08 13:48:49","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":10169064,"visible":true,"origin":"","legend":"Article File","description":"","filename":"CrossDNA.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8332638/v1_covered_d4e0cb38-3e85-46bd-afe6-a1f77a57be35.pdf"},{"id":99757308,"identity":"879129bd-eb62-4bf8-ae57-0a3aa4fa425a","added_by":"auto","created_at":"2026-01-08 06:02:03","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":404721,"visible":true,"origin":"","legend":"Supplementary information","description":"","filename":"Supplementary.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8332638/v1/d46b2cf2b2cb7326861cd40e.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Explicit Dynamic Cross-Strand Interactions for DNA Sequence Language Modeling","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"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":"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-8332638/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8332638/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Large language models are driving a transformation in the functional interpretation of genomic sequences. Methodologically, existing DNA sequence language models can be broadly categorized into two types: one that treats DNA sequence as unidirectional text for single-strand modeling, and another that achieves reverse-complement symmetry through data augmentation or model equivariance for static double-strand modeling. Both approaches predominantly approximate double-strand interactions in an implicit and static manner, struggling to capture context-driven cross-strand information exchange during sequence representation learning. In reality, double-strand information exchange is not an isolated event but is regulated by continuous physical coupling, functional synergy, and information transfer—a mechanism fundamental to genomic function. Based on this, we propose CrossDNA, an explicit and dynamic language model for DNA cross-strand modeling. Specifically, CrossDNA employs a dual-branch architecture with rotating input of double-strand sequence data to simulate the continuous information flow in the DNA double helix, establishes inter-strand communication via a lightweight TokenBridge module, and incorporates Comba with window-sliding attention (SWA) to capture long-range dependencies, while maintaining reverse-complement equivalence and stabilizing single-strand contextual semantics through self-distillation and consistency constraints from a branch teacher model. On tasks such as classification, regression, and representation, CrossDNA achieves consistent performance improvements and significantly enhances model robustness to sequence orientation, particularly in enhancer prediction where it more readily identifies features with clear biological significance. On multiple benchmarks we evaluated, CrossDNA, with a model size of only a few million parameters, matches or surpasses the performance of large models with hundreds of millions of parameters, substantially reducing training and inference costs and demonstrating high parameter efficiency and usability. Overall, CrossDNA advances DNA representation from implicit, static approximation to explicit, dynamic systematic modeling, signaling the orientation for a new generation of DNA language models and laying the foundation for deeper analysis of genomic structure and function.","manuscriptTitle":"Explicit Dynamic Cross-Strand Interactions for DNA Sequence Language Modeling","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-08 06:01:58","doi":"10.21203/rs.3.rs-8332638/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-machine-intelligence","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"natmachintell","sideBox":"Learn more about [Nature Machine Intelligence](http://www.nature.com/natmachintell/)","snPcode":"","submissionUrl":"","title":"Nature Machine Intelligence","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"c88fdec1-bfaf-457c-bb3b-da52bf1f81a3","owner":[],"postedDate":"January 8th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":60681650,"name":"Biological sciences/Computational biology and bioinformatics/Computational models"},{"id":60681651,"name":"Biological sciences/Biological techniques/Bioinformatics"}],"tags":[],"updatedAt":"2026-05-05T14:50:23+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-08 06:01:58","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8332638","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8332638","identity":"rs-8332638","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}