{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,7,28]],"date-time":"2026-07-28T23:46:20Z","timestamp":1785282380987,"version":"3.55.0"},"posted":{"date-parts":[[2025,5,5]]},"group-title":"Psychiatry and Clinical Psychology","reference-count":69,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2025,5,5]],"date-time":"2025-05-05T00:00:00Z","timestamp":1746403200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2025,5,5]]},"abstract":"<jats:title>ABSTRACT<\/jats:title>\n                <jats:p>\n                  Polygenic scores (PGSs) have emerged as promising tools for predicting complex traits from genetic data, however, their predictive performance for psychiatric disorders remains limited and the added value of deep learning (DL) over linear models is underexplored. In this study, we compared our DL model, Genome-Local-Net (GLN), with the linear model bigstatsr in predicting five psychiatric disorders\u2014ADHD, ASD, BIP, MDD, and SCZ\u2014using individual-level genotype data. We further assessed whether combining these\n                  <jats:italic>internal<\/jats:italic>\n                  (individual-based) PGSs with\n                  <jats:italic>external<\/jats:italic>\n                  (GWAS-derived) PGSs and family genetic risk scores (FGRSs) could improve prediction additively or synergistically. While GLN and bigstatsr performed similarly in-sample, GLN showed better generalization on an out-of-sample replication set for ADHD, ASD, and MDD, with an average AUROC gain of 0.026. Integrating\n                  <jats:italic>internal<\/jats:italic>\n                  ,\n                  <jats:italic>external<\/jats:italic>\n                  , and family-based scores significantly improved ADHD prediction, though DL-based integration provided no consistent advantage over logistic models. These findings suggest that while DL may enhance generalizability for specific psychiatric traits, linear models remain competitive and effective for genetic risk prediction.\n                <\/jats:p>","DOI":"10.1101\/2025.05.05.25326794","type":"posted-content","created":{"date-parts":[[2025,5,5]],"date-time":"2025-05-05T20:05:11Z","timestamp":1746475511000},"source":"Crossref","is-referenced-by-count":2,"title":["Deep learning-based polygenic scores enhance generalizability of psychiatric disorders prediction"],"prefix":"10.64898","author":[{"given":"Leonardo","family":"Cobuccio","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Arnor I.","family":"Sigurdsson","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kajsa-Lotta","family":"Georgii Hellberg","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Morten","family":"Dybdahl Krebs","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jonas","family":"Meisner","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"name":"iPSYCH Study Consortium","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Thomas","family":"Werge","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael E.","family":"Benros","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Andrew J.","family":"Schork","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Simon","family":"Rasmussen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2025050707451171000_2025.05.05.25326794v1.1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ajhg.2021.04.014"},{"key":"2025050707451171000_2025.05.05.25326794v1.2","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-023-40330-w"},{"key":"2025050707451171000_2025.05.05.25326794v1.3","doi-asserted-by":"publisher","DOI":"10.1001\/jamapsychiatry.2022.4076"},{"key":"2025050707451171000_2025.05.05.25326794v1.4","doi-asserted-by":"publisher","DOI":"10.1002\/wps.21034"},{"key":"2025050707451171000_2025.05.05.25326794v1.5","doi-asserted-by":"publisher","DOI":"10.1038\/s10038-020-00832-7"},{"key":"2025050707451171000_2025.05.05.25326794v1.6","doi-asserted-by":"publisher","DOI":"10.1111\/j.2517-6161.1995.tb02031.x"},{"key":"2025050707451171000_2025.05.05.25326794v1.7","doi-asserted-by":"publisher","DOI":"10.1038\/s41380-020-0825-2"},{"key":"2025050707451171000_2025.05.05.25326794v1.8","doi-asserted-by":"publisher","DOI":"10.1101\/2020.11.30.20237768"},{"key":"2025050707451171000_2025.05.05.25326794v1.9","doi-asserted-by":"publisher","DOI":"10.1186\/s13742-015-0047-8"},{"key":"2025050707451171000_2025.05.05.25326794v1.10","doi-asserted-by":"publisher","DOI":"10.1038\/s41596-020-0353-1"},{"key":"2025050707451171000_2025.05.05.25326794v1.11","doi-asserted-by":"publisher","DOI":"10.1093\/gigascience\/giz082"},{"key":"2025050707451171000_2025.05.05.25326794v1.12","doi-asserted-by":"publisher","DOI":"10.1016\/j.jclinepi.2019.02.004"},{"key":"2025050707451171000_2025.05.05.25326794v1.13","doi-asserted-by":"publisher","DOI":"10.1093\/gigascience\/giab008"},{"key":"2025050707451171000_2025.05.05.25326794v1.14","doi-asserted-by":"publisher","DOI":"10.1038\/s41588-022-01285-8"},{"key":"2025050707451171000_2025.05.05.25326794v1.15","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-023-06079-4"},{"key":"2025050707451171000_2025.05.05.25326794v1.16","doi-asserted-by":"publisher","DOI":"10.1101\/2023.06.19.23291606"},{"key":"2025050707451171000_2025.05.05.25326794v1.17","doi-asserted-by":"publisher","DOI":"10.1016\/j.ajhg.2024.09.009"},{"key":"2025050707451171000_2025.05.05.25326794v1.18","doi-asserted-by":"publisher","DOI":"10.1038\/s42003-022-03812-z"},{"key":"2025050707451171000_2025.05.05.25326794v1.19","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-019-09718-5"},{"key":"2025050707451171000_2025.05.05.25326794v1.20","doi-asserted-by":"publisher","DOI":"10.1101\/2024.10.25.24316156"},{"key":"2025050707451171000_2025.05.05.25326794v1.21","doi-asserted-by":"publisher","DOI":"10.1038\/s41588-019-0344-8"},{"key":"2025050707451171000_2025.05.05.25326794v1.22","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-021-03765-z"},{"key":"2025050707451171000_2025.05.05.25326794v1.23","doi-asserted-by":"publisher","DOI":"10.1101\/2023.02.10.23285766"},{"key":"2025050707451171000_2025.05.05.25326794v1.24","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pgen.1000008"},{"key":"2025050707451171000_2025.05.05.25326794v1.25","doi-asserted-by":"publisher","DOI":"10.1016\/j.ajhg.2021.02.014"},{"key":"2025050707451171000_2025.05.05.25326794v1.26","doi-asserted-by":"publisher","DOI":"10.1038\/s41593-018-0326-7"},{"key":"2025050707451171000_2025.05.05.25326794v1.27","doi-asserted-by":"publisher","DOI":"10.1016\/j.xgen.2022.100152"},{"key":"2025050707451171000_2025.05.05.25326794v1.28","doi-asserted-by":"publisher","DOI":"10.1001\/jamapsychiatry.2021.0336"},{"key":"2025050707451171000_2025.05.05.25326794v1.29","doi-asserted-by":"publisher","DOI":"10.1038\/s41588-018-0183-z"},{"key":"2025050707451171000_2025.05.05.25326794v1.30","doi-asserted-by":"crossref","unstructured":"Klau, J. H. , Maj, C. , Klinkhammer, H. , Krawitz, P. M. , Mayr, A. , Hillmer, A. M. , Schumacher, J. , & Heider, D . (2023). AI-based multi-PRS models outperform classical single-PRS models. Frontiers in Genetics, 14. https:\/\/www.frontiersin.org\/articles\/10.3389\/fgene.2023.1217860","DOI":"10.3389\/fgene.2023.1217860"},{"key":"2025050707451171000_2025.05.05.25326794v1.31","doi-asserted-by":"publisher","DOI":"10.1038\/mp.2017.163"},{"key":"2025050707451171000_2025.05.05.25326794v1.32","doi-asserted-by":"publisher","DOI":"10.1038\/s41576-022-00470-z"},{"key":"2025050707451171000_2025.05.05.25326794v1.33","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btab452"},{"key":"2025050707451171000_2025.05.05.25326794v1.34","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-019-12653-0"},{"key":"2025050707451171000_2025.05.05.25326794v1.35","doi-asserted-by":"publisher","DOI":"10.1038\/ng.3190"},{"key":"2025050707451171000_2025.05.05.25326794v1.36","doi-asserted-by":"publisher","DOI":"10.1007\/s00125-023-05955-y"},{"key":"2025050707451171000_2025.05.05.25326794v1.37","doi-asserted-by":"publisher","DOI":"10.1177\/1403494811401482"},{"key":"2025050707451171000_2025.05.05.25326794v1.38","doi-asserted-by":"publisher","DOI":"10.1038\/s42256-024-00942-3"},{"key":"2025050707451171000_2025.05.05.25326794v1.39","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-020-0800-0"},{"key":"2025050707451171000_2025.05.05.25326794v1.40","doi-asserted-by":"publisher","DOI":"10.1016\/j.ajhg.2022.10.009"},{"key":"2025050707451171000_2025.05.05.25326794v1.41","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-018-30089-2"},{"key":"2025050707451171000_2025.05.05.25326794v1.42","doi-asserted-by":"publisher","DOI":"10.1177\/1403494810395825"},{"key":"2025050707451171000_2025.05.05.25326794v1.43","doi-asserted-by":"publisher","DOI":"10.1038\/s41588-021-00857-4"},{"key":"2025050707451171000_2025.05.05.25326794v1.44","doi-asserted-by":"publisher","DOI":"10.1001\/jamapsychiatry.2020.3042"},{"key":"2025050707451171000_2025.05.05.25326794v1.45","doi-asserted-by":"publisher","DOI":"10.1007\/s10545-007-0631-x"},{"key":"2025050707451171000_2025.05.05.25326794v1.46","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-024-48654-x"},{"key":"2025050707451171000_2025.05.05.25326794v1.47","doi-asserted-by":"publisher","DOI":"10.1038\/mp.2017.196"},{"key":"2025050707451171000_2025.05.05.25326794v1.48","doi-asserted-by":"publisher","DOI":"10.1016\/j.fmre.2024.02.015"},{"key":"2025050707451171000_2025.05.05.25326794v1.49","doi-asserted-by":"publisher","DOI":"10.1101\/cshperspect.a040493"},{"key":"2025050707451171000_2025.05.05.25326794v1.50","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btaa1029"},{"key":"2025050707451171000_2025.05.05.25326794v1.51","doi-asserted-by":"publisher","DOI":"10.1534\/genetics.119.302019"},{"key":"2025050707451171000_2025.05.05.25326794v1.52","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/bty185"},{"key":"2025050707451171000_2025.05.05.25326794v1.53","doi-asserted-by":"publisher","DOI":"10.1101\/2020.10.30.362079"},{"key":"2025050707451171000_2025.05.05.25326794v1.54","doi-asserted-by":"publisher","DOI":"10.1038\/nature08185"},{"key":"2025050707451171000_2025.05.05.25326794v1.55","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pgen.1009141"},{"key":"2025050707451171000_2025.05.05.25326794v1.56","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1811.12808"},{"key":"2025050707451171000_2025.05.05.25326794v1.57","doi-asserted-by":"publisher","DOI":"10.1038\/s41588-018-0213-x"},{"key":"2025050707451171000_2025.05.05.25326794v1.58","doi-asserted-by":"publisher","DOI":"10.1038\/s41593-018-0320-0"},{"key":"2025050707451171000_2025.05.05.25326794v1.59","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbae240"},{"key":"2025050707451171000_2025.05.05.25326794v1.60","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1922305118"},{"key":"2025050707451171000_2025.05.05.25326794v1.61","doi-asserted-by":"publisher","DOI":"10.1101\/2024.07.04.24309942"},{"key":"2025050707451171000_2025.05.05.25326794v1.62","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkad373"},{"key":"2025050707451171000_2025.05.05.25326794v1.63","doi-asserted-by":"publisher","DOI":"10.1101\/2022.10.27.22281549"},{"key":"2025050707451171000_2025.05.05.25326794v1.64","doi-asserted-by":"publisher","DOI":"10.1101\/2022.11.29.518075"},{"key":"2025050707451171000_2025.05.05.25326794v1.65","doi-asserted-by":"publisher","DOI":"10.1038\/s41576-018-0018-x"},{"key":"2025050707451171000_2025.05.05.25326794v1.66","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-022-04434-5"},{"key":"2025050707451171000_2025.05.05.25326794v1.67","doi-asserted-by":"publisher","DOI":"10.1016\/j.xgen.2024.100523"},{"key":"2025050707451171000_2025.05.05.25326794v1.68","doi-asserted-by":"publisher","DOI":"10.1146\/annurev-biodatasci-111721-074830"},{"key":"2025050707451171000_2025.05.05.25326794v1.69","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pgen.1009697"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1101\/2025.05.05.25326794","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T20:35:25Z","timestamp":1768509325000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.1101\/2025.05.05.25326794"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,5,5]]},"references-count":69,"URL":"https:\/\/doi.org\/10.1101\/2025.05.05.25326794","relation":{},"subject":[],"published":{"date-parts":[[2025,5,5]]},"subtype":"preprint"}}