{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,22]],"date-time":"2026-10-22T10:40:22Z","timestamp":1792665622898,"version":"4.5.0"},"reference-count":28,"publisher":"Oxford University Press (OUP)","issue":"5","license":[{"start":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T00:00:00Z","timestamp":1792022400000},"content-version":"vor","delay-in-days":44,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100003725","name":"National Research Foundation of Korea","doi-asserted-by":"publisher","award":["2020H1D3A2A03100666"],"award-info":[{"award-number":["2020H1D3A2A03100666"]}],"id":[{"id":"10.13039\/501100003725","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100003569","name":"Ministry of Food and Drug Safety","doi-asserted-by":"publisher","award":["RS-2026-25506298"],"award-info":[{"award-number":["RS-2026-25506298"]}],"id":[{"id":"10.13039\/501100003569","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,9,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Polygenic risk scores (PRSs) are widely used to quantify genetic predisposition to complex traits, and their predictive accuracy can be enhanced by incorporating functional annotations into single-nucleotide polymorphism (SNP) effect-size estimation. Existing approaches typically assume linear relationships between annotations and SNP effects, neglecting predictive aleatoric variance and potential nonlinear regulatory interactions. To evaluate the gains of nonlinear and uncertainty-aware modeling, we developed DeepU-PRS, a deep ensemble framework that extends the linear BLD-LDAK model with a neural network trained on functional annotations to jointly estimate SNP heritability and its aleatoric uncertainty, which are then integrated into a Bayesian PRS model. Across 28 continuous and 8 disease phenotypes in the UK Biobank, DeepU-PRS improved the mean continuous R2 by approximately 10% over linear BLD-LDAK, with the largest gains for traits whose association signals were concentrated, whereas the more diffuse disease traits favored BLD-LDAK. These patterns were reproduced in individual-level data from All of Us. Against a panel of established methods (C\u2009+\u2009T, Lassosum, PRS-CS, LDpred-funct, and SBayesRC), DeepU-PRS was among the strongest single methods on continuous traits, and ensembles combining DeepU-PRS with the linear baseline and SBayesRC further improved accuracy, indicating complementary predictive signal. These findings highlight the value of modeling nonlinear and uncertain regulatory effects to advance PRS prediction.<\/jats:p>","DOI":"10.1093\/bib\/bbag563","type":"journal-article","created":{"date-parts":[[2026,10,7]],"date-time":"2026-10-07T11:51:34Z","timestamp":1791373894000},"source":"Crossref","is-referenced-by-count":0,"title":["DeepU-PRS: nonlinear and uncertainty-aware SNP heritability modeling for annotation-informed polygenic risk scoring"],"prefix":"10.1093","volume":"27","author":[{"given":"Junhyeong","family":"Lee","sequence":"first","affiliation":[{"name":"Graduate School of Data Science, Seoul National University , Seoul ,","place":["South Korea"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4592-9564","authenticated-orcid":false,"given":"Dokyoon","family":"Kim","sequence":"additional","affiliation":[{"name":"Department of Biostatistics, Epidemiology & Informatics, The Perelman School of Medicine, University of Pennsylvania , Philadelphia, PA ,","place":["United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8097-3878","authenticated-orcid":false,"given":"Seunggeun","family":"Lee","sequence":"additional","affiliation":[{"name":"Graduate School of Data Science, Seoul National University , Seoul ,","place":["South Korea"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2026,10,15]]},"reference":[{"key":"2026102206065903800_ref1","doi-asserted-by":"publisher","first-page":"581","DOI":"10.1038\/s41576-018-0018-x","article-title":"The personal and clinical utility of polygenic risk scores","volume":"19","author":"Torkamani","year":"2018","journal-title":"Nat Rev Genet"},{"key":"2026102206065903800_ref2","doi-asserted-by":"publisher","first-page":"400","DOI":"10.1038\/ng.2579","article-title":"Projecting the performance of risk prediction based on polygenic analyses of GWAS","volume":"45","author":"Chatterjee","year":"2013","journal-title":"Nat Genet"},{"key":"2026102206065903800_ref3","doi-asserted-by":"publisher","first-page":"748","DOI":"10.1038\/nature08185","article-title":"Common polygenic variation contributes to risk of schizophrenia and bipolar disorder","volume":"460","author":"Purcell","year":"2009","journal-title":"Nature"},{"key":"2026102206065903800_ref4","doi-asserted-by":"publisher","first-page":"1228","DOI":"10.1038\/ng.3404","article-title":"Partitioning heritability by functional annotation using genome-wide association summary statistics","volume":"47","author":"Finucane","year":"2015","journal-title":"Nat Genet"},{"key":"2026102206065903800_ref5","doi-asserted-by":"publisher","first-page":"277","DOI":"10.1038\/s41588-018-0279-5","article-title":"SumHer better estimates the SNP heritability of complex traits from summary statistics","volume":"51","author":"Speed","year":"2019","journal-title":"Nat Genet"},{"key":"2026102206065903800_ref6","doi-asserted-by":"publisher","first-page":"1236","DOI":"10.1038\/ng.3406","article-title":"An atlas of genetic correlations across human diseases and traits","volume":"47","author":"Bulik-Sullivan","year":"2015","journal-title":"Nat Genet"},{"key":"2026102206065903800_ref7","doi-asserted-by":"publisher","first-page":"4192","DOI":"10.1038\/s41467-021-24485-y","article-title":"Improved genetic prediction of complex traits from individual-level data or summary statistics","volume":"12","author":"Zhang","year":"2021","journal-title":"Nat Commun"},{"key":"2026102206065903800_ref8","volume-title":"Proceedings of the 33rd International Conference on Machine Learning","author":"Gal"},{"key":"2026102206065903800_ref9","first-page":"6402","article-title":"Simple and scalable predictive uncertainty estimation using deep ensembles","volume":"30","author":"Lakshminarayanan","year":"2017","journal-title":"Adv Neural Inf Process Syst"},{"key":"2026102206065903800_ref10","doi-asserted-by":"publisher","first-page":"469","DOI":"10.1002\/gepi.22050","article-title":"Polygenic scores via penalized regression on summary statistics","volume":"41","author":"Mak","year":"2017","journal-title":"Genet Epidemiol"},{"key":"2026102206065903800_ref11","doi-asserted-by":"publisher","first-page":"1776","DOI":"10.1038\/s41467-019-09718-5","article-title":"Polygenic prediction via Bayesian regression and continuous shrinkage priors","volume":"10","author":"Ge","year":"2019","journal-title":"Nat Commun"},{"key":"2026102206065903800_ref12","doi-asserted-by":"publisher","first-page":"6052","DOI":"10.1038\/s41467-021-25171-9","article-title":"Incorporating functional priors improves polygenic prediction accuracy in UK biobank and 23andMe data","volume":"12","author":"M\u00e1rquez-Luna","year":"2021","journal-title":"Nat Commun"},{"key":"2026102206065903800_ref13","doi-asserted-by":"publisher","first-page":"767","DOI":"10.1038\/s41588-024-01704-y","article-title":"Leveraging functional genomic annotations and genome coverage to improve polygenic prediction of complex traits within and between ancestries","volume":"56","author":"Zheng","year":"2024","journal-title":"Nat Genet"},{"key":"2026102206065903800_ref14","doi-asserted-by":"publisher","first-page":"100850","DOI":"10.1016\/j.xgen.2025.100850","article-title":"Incorporating multiple functional annotations to improve polygenic risk prediction accuracy","volume":"5","author":"Shao","year":"2025","journal-title":"Cell Genom"},{"key":"2026102206065903800_ref15","doi-asserted-by":"publisher","first-page":"e0307270","DOI":"10.1371\/journal.pone.0307270","article-title":"UK biobank release and systematic evaluation of optimised polygenic risk scores for 53 diseases and quantitative traits","volume":"19","author":"Thompson","year":"2024","journal-title":"PLoS One"},{"key":"2026102206065903800_ref16","doi-asserted-by":"publisher","first-page":"1196","DOI":"10.1038\/s41592-021-01252-x","article-title":"Effective gene expression prediction from sequence by integrating long-range interactions","volume":"18","author":"Avsec","year":"2021","journal-title":"Nat Methods"},{"key":"2026102206065903800_ref17","doi-asserted-by":"publisher","first-page":"49","DOI":"10.1186\/s13059-020-02252-4","article-title":"Exploiting the GTEx resources to decipher the mechanisms at GWAS loci","volume":"22","author":"Barbeira","year":"2021","journal-title":"Genome Biol"},{"key":"2026102206065903800_ref18","doi-asserted-by":"publisher","first-page":"1091","DOI":"10.1038\/ng.3367","article-title":"A gene-based association method for mapping traits using reference transcriptome data","volume":"47","author":"Gamazon","year":"2015","journal-title":"Nat Genet"},{"key":"2026102206065903800_ref19","doi-asserted-by":"publisher","first-page":"1825","DOI":"10.1038\/s41467-018-03621-1","article-title":"Exploring the phenotypic consequences of tissue-specific gene expression variation inferred from GWAS summary statistics","volume":"9","author":"Barbeira","year":"2018","journal-title":"Nat Commun"},{"key":"2026102206065903800_ref20","doi-asserted-by":"publisher","first-page":"340","DOI":"10.1038\/s41586-023-06957-x","article-title":"Genomic data in the all of us research program","volume":"627","author":"All of Us Research Program Genomics Investigators","year":"2024","journal-title":"Nature"},{"key":"2026102206065903800_ref21","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1016\/j.neunet.2017.12.012","article-title":"Sigmoid-weighted linear units for neural network function approximation in reinforcement learning","volume":"107","author":"Elfwing","year":"2018","journal-title":"Neural Netw"},{"key":"2026102206065903800_ref22","first-page":"472","article-title":"Incorporating second-order functional knowledge for better option pricing","volume":"13","author":"Dugas","year":"2001","journal-title":"Adv Neural Inf Process Syst"},{"key":"2026102206065903800_ref23","article-title":"Why gradient clipping accelerates training: a theoretical justification for adaptivity","author":"Zhang","journal-title":"Proceedings of the 8th International Conference on Learning Representations"},{"key":"2026102206065903800_ref24","article-title":"Large batch optimization for deep learning: training BERT in 76 minutes","author":"You","journal-title":"Proceedings of the 8th International Conference on Learning Representations"},{"key":"2026102206065903800_ref25","author":"PyTorch"},{"key":"2026102206065903800_ref26","doi-asserted-by":"publisher","first-page":"5086","DOI":"10.1038\/s41467-019-12653-0","article-title":"Improved polygenic prediction by Bayesian multiple regression on summary statistics","volume":"10","author":"Lloyd-Jones","year":"2019","journal-title":"Nat Commun"},{"key":"2026102206065903800_ref27","author":"Speed"},{"key":"2026102206065903800_ref28","doi-asserted-by":"publisher","first-page":"68","DOI":"10.1038\/nature15393","article-title":"A global reference for human genetic variation","volume":"526","author":"1000 Genomes Project Consortium","year":"2015","journal-title":"Nature"}],"container-title":["Briefings in Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bib\/article-pdf\/27\/5\/bbag563\/71972816\/bbag563.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bib\/article-pdf\/27\/5\/bbag563\/71972816\/bbag563.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,22]],"date-time":"2026-10-22T10:07:06Z","timestamp":1792663626000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bib\/article\/doi\/10.1093\/bib\/bbag563\/8918435"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,1]]},"references-count":28,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2026,9,1]]}},"URL":"https:\/\/doi.org\/10.1093\/bib\/bbag563","relation":{},"ISSN":["1467-5463","1477-4054"],"issn-type":[{"value":"1467-5463","type":"print"},{"value":"1477-4054","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2026,9]]},"published":{"date-parts":[[2026,9,1]]},"article-number":"bbag563"}}