Multilingual transfer ability: Find Rosetta Stone between DNA Language and Natural Language

preprint OA: closed CC-BY-4.0
📄 Open PDF Full text JSON View at publisher

Abstract

Abstract This study aims to explore whether Large Language Models (LLMs) can transfer abstract structural reasoning capabilities from natural language to the genetic language, which lacks explicit semantics, thereby finding a "Rosetta Stone" to connect the two domains. We validated this hypothesis through a dual experimental design: first, a standard LLM fine-tuned on a natural language similarity task (PAWS-X) was used to assess biological sequence similarity; second, a custom model pre-trained on a multimodal corpus (including natural language, DNA, and protein) was fine-tuned in the same manner to determine the correct alignment of DNA-protein coding pairs. The results show that the transfer of basic similarity judgment ability was successful (with accuracy up to 89%), while for the more complex coding alignment task, the multimodal pre-trained model achieved a zero-shot accuracy of 81%. This study confirms that abstract structural pattern recognition can be transferred between the two languages, with its effectiveness highly dependent on the structural similarity of the tasks, and that multimodal pre-training is key to enabling complex rule transfer, establishing a new paradigm for using LLMs in biological discovery.
Full text 17,654 characters · extracted from preprint-html · click to expand
Multilingual transfer ability: Find Rosetta Stone between DNA Language and Natural Language | 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 Multilingual transfer ability: Find Rosetta Stone between DNA Language and Natural Language Wang Liang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7898312/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This study aims to explore whether Large Language Models (LLMs) can transfer abstract structural reasoning capabilities from natural language to the genetic language, which lacks explicit semantics, thereby finding a "Rosetta Stone" to connect the two domains. We validated this hypothesis through a dual experimental design: first, a standard LLM fine-tuned on a natural language similarity task (PAWS-X) was used to assess biological sequence similarity; second, a custom model pre-trained on a multimodal corpus (including natural language, DNA, and protein) was fine-tuned in the same manner to determine the correct alignment of DNA-protein coding pairs. The results show that the transfer of basic similarity judgment ability was successful (with accuracy up to 89%), while for the more complex coding alignment task, the multimodal pre-trained model achieved a zero-shot accuracy of 81%. This study confirms that abstract structural pattern recognition can be transferred between the two languages, with its effectiveness highly dependent on the structural similarity of the tasks, and that multimodal pre-training is key to enabling complex rule transfer, establishing a new paradigm for using LLMs in biological discovery. Biological sciences/Computational biology and bioinformatics Physical sciences/Mathematics and computing Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted 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-7898312","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":542552211,"identity":"d313a50b-5c02-494f-8632-fc8d16c5f36e","order_by":0,"name":"Wang Liang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA9UlEQVRIiWNgGAWjYLCCBwYgkvHhAyDJw0eUlgSwFmZjEMXDRpwWMMlsJgGiCGoxOH728IuEArs8eQdmtsqvOXYybAzMDx/dwKflTF6aRYJBcrHhAWa227LbkoEOYzM2zsGjxexAjplBggFz4sYG/mO3JbcxA7XwsEnj1XL+DUhLPVALM1ux5LZ6IrTcyDF+kGBwOHE+AzMb48dthwlrsb/xxgwYyMcTNzAzM0szbjvOw8ZMwC+S/TnGHz78qU6c397M+PHntmp7fvbmh4/xaQECNnB0GBwGxgwPiMWMXzlYyQcQKd8ATDI/CKseBaNgFIyCEQgAZaJEahMEgeAAAAAASUVORK5CYII=","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Wang","middleName":"","lastName":"Liang","suffix":""}],"badges":[],"createdAt":"2025-10-19 11:53:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7898312/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7898312/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":96215058,"identity":"d8d3387a-7c0f-4736-ad2c-3e707b1dff0e","added_by":"auto","created_at":"2025-11-18 20:01:59","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1018779,"visible":true,"origin":"","legend":"","description":"","filename":"MultilingualtransferabilityFindRosettaStonebetweenDNALanguageandNaturalLanguage.docx","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/1534c8ebf98d618d9eede7e8.docx"},{"id":96215050,"identity":"db07b9fe-aee7-49dc-b29f-6b811ecb5274","added_by":"auto","created_at":"2025-11-18 20:01:59","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3300,"visible":true,"origin":"","legend":"","description":"","filename":"4120d8a2953d4f83864f27334f8be067.json","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/267a695c0a6c551e80e9ae3c.json"},{"id":96253150,"identity":"fb8222c9-90c3-4c9b-8d84-9004bc631201","added_by":"auto","created_at":"2025-11-19 07:42:02","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":151182,"visible":true,"origin":"","legend":"","description":"","filename":"4120d8a2953d4f83864f27334f8be0671enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/cf565a301ab46f8b9ec2ccb5.xml"},{"id":96215053,"identity":"df66bd59-f419-4134-851b-6c8b6a643b45","added_by":"auto","created_at":"2025-11-18 20:01:59","extension":"jpeg","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":329225,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/4355cf10fa6ef40b6482f5d8.jpeg"},{"id":96215051,"identity":"32cc578d-d547-47d5-9ed4-d6fa825bce74","added_by":"auto","created_at":"2025-11-18 20:01:59","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":38729,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/969fe1318012d401208922ed.png"},{"id":96215063,"identity":"a2fa0658-e3d5-4687-93ac-d274a6d72f26","added_by":"auto","created_at":"2025-11-18 20:02:00","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":221777,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/2b084f0297fbfde07ad95a2c.png"},{"id":96251036,"identity":"6ae371eb-42bd-43c1-9552-269c9fd96be7","added_by":"auto","created_at":"2025-11-19 07:39:15","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":277569,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/96e93691b14b362d3698403e.jpeg"},{"id":96251799,"identity":"0fc86654-4159-4f6d-8312-282d2ee00840","added_by":"auto","created_at":"2025-11-19 07:40:03","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":21885,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/3a3776884d713f8b476a54a2.png"},{"id":96251811,"identity":"89313609-2565-4350-9ead-bafe79195f4a","added_by":"auto","created_at":"2025-11-19 07:40:04","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42236,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/8a98f0ae482bac3e2d131c6a.png"},{"id":96252641,"identity":"083af505-597d-4c8f-92eb-6d281008424b","added_by":"auto","created_at":"2025-11-19 07:41:18","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":147340,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/c6b6971bb256d4b8a87510b1.png"},{"id":96215055,"identity":"6effa630-49af-43b4-9505-9aad46c77c14","added_by":"auto","created_at":"2025-11-18 20:01:59","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":69121,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/73803703136394b3257deb0f.png"},{"id":96215052,"identity":"70eb54ee-91cb-46af-9714-859ed7bbc09f","added_by":"auto","created_at":"2025-11-18 20:01:59","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13672,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/2d962e2bdfe43b7d041027ff.png"},{"id":96215060,"identity":"8274e994-69e4-491a-91b2-c85645618816","added_by":"auto","created_at":"2025-11-18 20:01:59","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53061,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/f40d79934bde431a627c8433.png"},{"id":96215065,"identity":"c1fc0461-adb4-481f-9cd2-3cd79491c9f5","added_by":"auto","created_at":"2025-11-18 20:02:00","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":69228,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/5ad3409fac11249b11afbebe.png"},{"id":96215062,"identity":"9bcd750b-1e89-449b-8cae-086dfac1187b","added_by":"auto","created_at":"2025-11-18 20:02:00","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9244,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/541f33a1b2af6e34a912a376.png"},{"id":96250897,"identity":"9fdcff5c-101d-4974-9f6e-19864b2aa353","added_by":"auto","created_at":"2025-11-19 07:39:06","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12681,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/ef41c085cf4a82c109d97f20.png"},{"id":96215067,"identity":"5c2bc2e5-b001-4349-94c2-0032c3009a87","added_by":"auto","created_at":"2025-11-18 20:02:00","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42412,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/9cbc80a4732dee2bef09bdb4.png"},{"id":96252941,"identity":"3cd6a28c-c64b-44c4-8e24-26a73c6e5574","added_by":"auto","created_at":"2025-11-19 07:41:41","extension":"xml","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":146799,"visible":true,"origin":"","legend":"","description":"","filename":"4120d8a2953d4f83864f27334f8be0671structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/42e6f3cf1e8e103f6894a53b.xml"},{"id":96215068,"identity":"795f1946-5ec5-4259-a01e-9006e7ae0f3d","added_by":"auto","created_at":"2025-11-18 20:02:00","extension":"html","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":165578,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1/08eec4081f3f188ae5f7b848.html"},{"id":98629639,"identity":"5a4d95fe-8fce-4d74-8c85-0f8fee112645","added_by":"auto","created_at":"2025-12-19 17:14:24","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":848219,"visible":true,"origin":"","legend":"","description":"","filename":"MultilingualtransferabilityFindRosettaStonebetweenDNALanguageandNaturalLanguage.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7898312/v1_covered_6215a745-4a3f-4857-b914-8b5f77d6e232.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Multilingual transfer ability: Find Rosetta Stone between DNA Language and Natural Language","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"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":"","lastPublishedDoi":"10.21203/rs.3.rs-7898312/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7898312/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study aims to explore whether Large Language Models (LLMs) can transfer abstract structural reasoning capabilities from natural language to the genetic language, which lacks explicit semantics, thereby finding a \"Rosetta Stone\" to connect the two domains. We validated this hypothesis through a dual experimental design: first, a standard LLM fine-tuned on a natural language similarity task (PAWS-X) was used to assess biological sequence similarity; second, a custom model pre-trained on a multimodal corpus (including natural language, DNA, and protein) was fine-tuned in the same manner to determine the correct alignment of DNA-protein coding pairs. The results show that the transfer of basic similarity judgment ability was successful (with accuracy up to 89%), while for the more complex coding alignment task, the multimodal pre-trained model achieved a zero-shot accuracy of 81%. This study confirms that abstract structural pattern recognition can be transferred between the two languages, with its effectiveness highly dependent on the structural similarity of the tasks, and that multimodal pre-training is key to enabling complex rule transfer, establishing a new paradigm for using LLMs in biological discovery.\u003c/p\u003e","manuscriptTitle":"Multilingual transfer ability: Find Rosetta Stone between DNA Language and Natural Language","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-18 20:01:55","doi":"10.21203/rs.3.rs-7898312/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":"3373c7bb-6aab-4ead-9ac6-f3fc11e0d7fc","owner":[],"postedDate":"November 18th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":57714893,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":57714894,"name":"Physical sciences/Mathematics and computing"}],"tags":[],"updatedAt":"2025-12-19T15:39:23+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-18 20:01:55","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7898312","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7898312","identity":"rs-7898312","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","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.

My notes (saved in your browser only)

Ask this paper AI returns verbatim quotes from the full text · source: preprint-html

Answers must be backed by verbatim quotes from this paper's full text. Hallucinated quotes are dropped automatically; if no verbatim passage answers the question, we say so. How this works

Citation neighborhood (no data yet)

We don't have any in-corpus citations linked to this paper yet. This is a recent paper (2025) — citers typically take a year or two to land, and the OpenAlex reference graph may still be filling in.

Source provenance

europepmc
last seen: 2026-05-20T01:45:00.602351+00:00
unpaywall
last seen: 2026-05-30T02:00:01.510937+00:00
License: CC-BY-4.0