Improving Polygenic Risk Score Based Drug Response Prediction Using Transfer Learning

Improving Polygenic Risk Score Based Drug Response Prediction Using Transfer Learning | 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 Improving Polygenic Risk Score Based Drug Response Prediction Using Transfer Learning Youshu Cheng, Song Zhai, Wujuan Zhong, Rachel Marceau West, Judong Shen This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6173450/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 21 Nov, 2025 Read the published version in npj Genomic Medicine → Version 1 posted 11 You are reading this latest preprint version Abstract Pharmacogenomics (PGx) studies aim to perform drug response prediction and patient stratification using genome-wide association study (GWAS) data from randomized clinical trials. Polygenic risk scores (PRS) are useful tools for PGx. By combining information across the genome, they have shown great promise in predicting disease risk and how patients respond to a particular treatment. A common practice when developing polygenic models for drug response prediction, is to use disease GWAS summary statistics derived from large cohorts of related disease phenotypes. However, this disease PRS approach (PRS-Dis) lacks the ability to incorporate any predictive (or genotype-by-treatment interaction) effects in the PRS training stage and thus cannot fully capture the heritability of drug response, often resulting in poor predictive performance. On the other hand, a direct PGx PRS approach (PRS-PGx) requires an independent PGx GWAS dataset with the same or similar drug response phenotype, which is usually not available. To fill this gap, we propose a transfer learning (TL) based method (PRS-PGx-TL) that jointly models large-scale disease GWAS summary statistics from the base (training) cohort and individual-level PGx data from the target cohort, leveraging both for parameter optimization and prognostic and predictive PRS construction. In PRS-PGx-TL, we develop a two-dimensional penalized gradient descent algorithm, which utilizes the PRS weights from the disease GWAS as initial values and optimizes the tuning parameters using a cross-validation framework while updating both prognostic and predictive effect estimates simultaneously. Through extensive simulation studies, we show that PRS-PGx-TL improves prediction accuracy and population stratification performance compared to the traditional PRS-Dis methods (e.g., PRS-CS, Lassosum). We further demonstrate its advantages by applying it to the IMPROVE-IT PGx GWAS data for predicting treatment related LDL cholesterol reduction. Overall, our proposed TL-based PRS method shows great value in improving drug response prediction and patient stratification and can help facilitate precision medicine by using an individual’s genotype information to guide treatment. Biological sciences/Computational biology and bioinformatics Biological sciences/Drug discovery Biological sciences/Genetics Health sciences/Biomarkers Health sciences/Diseases Health sciences/Medical research Health sciences/Molecular medicine Health sciences/Risk factors GWAS Pharmacogenomics Polygenic risk score Transfer learning Two-dimensional penalized gradient descent algorithm Full Text Additional Declarations Competing interest reported. S.Z., W.Z., R.M.W and J.S. are employees at Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA. Y.C. declares no competing interests. Supplementary Files PRSPGxTLsupp.docx Cite Share Download PDF Status: Published Journal Publication published 21 Nov, 2025 Read the published version in npj Genomic Medicine → Version 1 posted Editorial decision: Revision requested 11 Jul, 2025 Reviews received at journal 07 Jul, 2025 Reviewers agreed at journal 22 May, 2025 Reviews received at journal 12 May, 2025 Reviews received at journal 08 May, 2025 Reviewers agreed at journal 14 Apr, 2025 Reviewers agreed at journal 13 Apr, 2025 Reviewers invited by journal 21 Mar, 2025 Editor assigned by journal 13 Mar, 2025 Submission checks completed at journal 11 Mar, 2025 First submitted to journal 06 Mar, 2025 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-6173450","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":427106300,"identity":"d581460f-18de-4bf7-9e3a-eeddd9510561","order_by":0,"name":"Youshu Cheng","email":"","orcid":"","institution":"Yale University","correspondingAuthor":false,"prefix":"","firstName":"Youshu","middleName":"","lastName":"Cheng","suffix":""},{"id":427106301,"identity":"1874b97c-f77f-4727-beb8-41d56f3f672e","order_by":1,"name":"Song Zhai","email":"","orcid":"","institution":"Merck \u0026 Co., Inc","correspondingAuthor":false,"prefix":"","firstName":"Song","middleName":"","lastName":"Zhai","suffix":""},{"id":427106306,"identity":"89f5902a-f8f8-448d-94e6-65421e3d22f6","order_by":2,"name":"Wujuan Zhong","email":"","orcid":"","institution":"Merck \u0026 Co., Inc","correspondingAuthor":false,"prefix":"","firstName":"Wujuan","middleName":"","lastName":"Zhong","suffix":""},{"id":427106307,"identity":"ecd038de-0518-44de-87d4-d625ca448ea2","order_by":3,"name":"Rachel Marceau West","email":"","orcid":"","institution":"Merck \u0026 Co., Inc","correspondingAuthor":false,"prefix":"","firstName":"Rachel","middleName":"Marceau","lastName":"West","suffix":""},{"id":427106309,"identity":"74e4837b-b512-439d-9912-8241856dfcde","order_by":4,"name":"Judong Shen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3UlEQVRIiWNgGAWjYBACxgYGAzADSDI+ADEkSNHCbECUFohiCMkmQZQW5hnJGx8X/GKwN2fvPVZ1o+KevGQD88NHN/A5bEZasfHMPobEnT3n0m7nnCk2nM3AZmycg1dLjpk0bw9DgsGNHLPbuW0JCXIMPGzSxGixN7j/xqw49x+xWnh+MDBuuMFjxpzbkJAgTVBLz7NiY94GicQNZ3KMpXOOJRjObCbgF8N2YIjx/LGxNzh+xvBzTk2CvMTx5oeP8WppAFnVhhwZzHiUg4A8mPxDQNUoGAWjYBSMbAAAVWFGPlZQKGEAAAAASUVORK5CYII=","orcid":"","institution":"Merck \u0026 Co., Inc","correspondingAuthor":true,"prefix":"","firstName":"Judong","middleName":"","lastName":"Shen","suffix":""}],"badges":[],"createdAt":"2025-03-06 21:08:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6173450/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6173450/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41525-025-00528-x","type":"published","date":"2025-11-21T15:58:11+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":96650370,"identity":"84be4acb-bddf-40c8-a738-c640cdec2891","added_by":"auto","created_at":"2025-11-24 16:11:36","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1196059,"visible":true,"origin":"","legend":"","description":"","filename":"PRSPGxTLms.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6173450/v1_covered_bd0a6508-ee18-40dd-b827-353e818fd47c.pdf"},{"id":78435259,"identity":"4770e8a4-5b04-4218-bf54-2466df8163ef","added_by":"auto","created_at":"2025-03-13 07:58:53","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":2176552,"visible":true,"origin":"","legend":"","description":"","filename":"PRSPGxTLsupp.docx","url":"https://assets-eu.researchsquare.com/files/rs-6173450/v1/7f0e1480aa6805351bd97efa.docx"}],"financialInterests":"Competing interest reported. S.Z., W.Z., R.M.W and J.S. are employees at Merck Sharp \u0026 Dohme LLC, a subsidiary of Merck \u0026 Co., Inc., Rahway, NJ, USA. Y.C. declares no competing interests.","formattedTitle":"Improving Polygenic Risk Score Based Drug Response Prediction Using Transfer Learning","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"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":"npj-genomic-medicine","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"npjgenmed","sideBox":"Learn more about [npj Genomic Medicine](http://www.nature.com/npjgenmed/)","snPcode":"41525","submissionUrl":"https://mts-npjgenmed.nature.com/","title":"npj Genomic Medicine","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"GWAS, Pharmacogenomics, Polygenic risk score, Transfer learning, Two-dimensional penalized gradient descent algorithm","lastPublishedDoi":"10.21203/rs.3.rs-6173450/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6173450/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePharmacogenomics (PGx) studies aim to perform drug response prediction and patient stratification using genome-wide association study (GWAS) data from randomized clinical trials. Polygenic risk scores (PRS) are useful tools for PGx. By combining information across the genome, they have shown great promise in predicting disease risk and how patients respond to a particular treatment. A common practice when developing polygenic models for drug response prediction, is to use disease GWAS summary statistics derived from large cohorts of related disease phenotypes. However, this disease PRS approach (PRS-Dis) lacks the ability to incorporate any predictive (or genotype-by-treatment interaction) effects in the PRS training stage and thus cannot fully capture the heritability of drug response, often resulting in poor predictive performance. On the other hand, a direct PGx PRS approach (PRS-PGx) requires an independent PGx GWAS dataset with the same or similar drug response phenotype, which is usually not available. To fill this gap, we propose a transfer learning (TL) based method (PRS-PGx-TL) that jointly models large-scale disease GWAS summary statistics from the base (training) cohort and individual-level PGx data from the target cohort, leveraging both for parameter optimization and prognostic and predictive PRS construction. In PRS-PGx-TL, we develop a two-dimensional penalized gradient descent algorithm, which utilizes the PRS weights from the disease GWAS as initial values and optimizes the tuning parameters using a cross-validation framework while updating both prognostic and predictive effect estimates simultaneously. Through extensive simulation studies, we show that PRS-PGx-TL improves prediction accuracy and population stratification performance compared to the traditional PRS-Dis methods (e.g., PRS-CS, Lassosum). We further demonstrate its advantages by applying it to the IMPROVE-IT PGx GWAS data for predicting treatment related LDL cholesterol reduction. Overall, our proposed TL-based PRS method shows great value in improving drug response prediction and patient stratification and can help facilitate precision medicine by using an individual\u0026rsquo;s genotype information to guide treatment.\u003c/p\u003e","manuscriptTitle":"Improving Polygenic Risk Score Based Drug Response Prediction Using Transfer Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-03-13 07:50:48","doi":"10.21203/rs.3.rs-6173450/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-07-12T01:44:56+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-07T21:07:09+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"272593501053051163799638225898018731413","date":"2025-05-23T03:49:37+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-12T09:57:26+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-08T15:59:14+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"144851778363857692716544216758211915417","date":"2025-04-14T15:30:34+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"13084033460429366841411468810252410249","date":"2025-04-14T00:18:26+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-03-21T12:14:27+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-03-13T20:22:08+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-03-11T08:32:12+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Genomic Medicine","date":"2025-03-06T21:01:43+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-genomic-medicine","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"npjgenmed","sideBox":"Learn more about [npj Genomic Medicine](http://www.nature.com/npjgenmed/)","snPcode":"41525","submissionUrl":"https://mts-npjgenmed.nature.com/","title":"npj Genomic Medicine","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"84f1f886-61a1-4417-af12-6ca814b82261","owner":[],"postedDate":"March 13th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":45509562,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":45509563,"name":"Biological sciences/Drug discovery"},{"id":45509564,"name":"Biological sciences/Genetics"},{"id":45509565,"name":"Health sciences/Biomarkers"},{"id":45509566,"name":"Health sciences/Diseases"},{"id":45509567,"name":"Health sciences/Medical research"},{"id":45509569,"name":"Health sciences/Molecular medicine"},{"id":45509571,"name":"Health sciences/Risk factors"}],"tags":[],"updatedAt":"2025-11-24T16:06:36+00:00","versionOfRecord":{"articleIdentity":"rs-6173450","link":"https://doi.org/10.1038/s41525-025-00528-x","journal":{"identity":"npj-genomic-medicine","isVorOnly":false,"title":"npj Genomic Medicine"},"publishedOn":"2025-11-21 15:58:11","publishedOnDateReadable":"November 21st, 2025"},"versionCreatedAt":"2025-03-13 07:50:48","video":"","vorDoi":"10.1038/s41525-025-00528-x","vorDoiUrl":"https://doi.org/10.1038/s41525-025-00528-x","workflowStages":[]},"version":"v1","identity":"rs-6173450","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6173450","identity":"rs-6173450","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}