{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"bioRxiv"}],"indexed":{"date-parts":[[2026,10,2]],"date-time":"2026-10-02T03:53:40Z","timestamp":1790913220202,"version":"4.1.0"},"posted":{"date-parts":[[2026,9,29]]},"group-title":"Bioinformatics","reference-count":34,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,9,29]],"date-time":"2026-09-29T00:00:00Z","timestamp":1790640000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"funder":[{"award":["U01AG070112"],"award-info":[{"award-number":["U01AG070112"]}],"id":[{"id":"https:\/\/ror.org\/049v75w11","id-type":"ROR","asserted-by":"publisher"}]},{"award":["R01AG081398"],"award-info":[{"award-number":["R01AG081398"]}],"id":[{"id":"https:\/\/ror.org\/049v75w11","id-type":"ROR","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,9,29]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>\n                  In the recent growing interest of AI research toward biology, genetic association studies of AI-derived phenotypes from high-content modalities such as images emerges as a powerful means for biological discovery. However, such AI-phenotyping methods still lacks a good optimization target and an efficient evaluation framework. The number of discovered loci was the major criterion for evaluating the quality of AI-derived phenotypes. However, the time and computational resources required for running the GWAS and subsequent loci-clumping are substantial, limiting the rapid development and iteration of deep learning representation algorithms. Here we present a 1000\u00d7 faster and lightweight framework,\n                  <jats:monospace>\n                    sum-h\n                    <jats:sup>2<\/jats:sup>\n                  <\/jats:monospace>\n                  , than traditional GWAS framework for evaluating genetic discovery through total heritability. We revisit tr(\n                  <jats:italic>P<\/jats:italic>\n                  <jats:sup>\u22121<\/jats:sup>\n                  <jats:italic>G<\/jats:italic>\n                  ), previously proposed in the context of evolutionary studies, as a measure of multi-phenotype heritability. We showed that\n                  <jats:monospace>\n                    sum-h\n                    <jats:sup>2<\/jats:sup>\n                  <\/jats:monospace>\n                  , the sum of heritability over phenotypic Principal Components (PCs), is equivalent to the linear transformation-invariant tr(\n                  <jats:italic>P<\/jats:italic>\n                  <jats:sup>\u22121<\/jats:sup>\n                  <jats:italic>G<\/jats:italic>\n                  ), through both theoretical proof and simulation studies. Moreover,\n                  <jats:monospace>\n                    sum-h\n                    <jats:sup>2<\/jats:sup>\n                  <\/jats:monospace>\n                  can be estimated rapidly with minimal information loss over a relatedness-enriched sample, while preserving the relative ranking of endophenotypes by GWAS loci counts. Based on selected UKB data and\n                  <jats:monospace>\n                    sum-h\n                    <jats:sup>2<\/jats:sup>\n                  <\/jats:monospace>\n                  , we set up a Genetic Discovery Arena, enabling rapid and fair comparisons for the development and optimization of deep learning-derived phenotyping methods.\n                <\/jats:p>","DOI":"10.64898\/2026.09.24.753959","type":"posted-content","created":{"date-parts":[[2026,9,30]],"date-time":"2026-09-30T00:15:25Z","timestamp":1790727325000},"source":"Crossref","is-referenced-by-count":0,"title":["Sum-h\n                  <sup>2<\/sup>\n                  , enabling genetic discovery for deep learning-derived phenotypes through a fast evaluation framework and arena of performance"],"prefix":"10.64898","author":[{"given":"Tian","family":"Xia","sequence":"first","affiliation":[{"name":"Department of Biomedical Informatics and Data Science, Yale School of Medicine, New Haven, CT, US"},{"name":"D. Bradley McWilliams School of Biomedical Informatics, The University of Texas Health Science Center at Houston, Houston, TX, US"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ziqian","family":"Xie","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics and Data Science, Yale School of Medicine, New Haven, CT, US"},{"name":"D. Bradley McWilliams School of Biomedical Informatics, The University of Texas Health Science Center at Houston, Houston, TX, US"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xingzhong","family":"Zhao","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics and Data Science, Yale School of Medicine, New Haven, CT, US"},{"name":"D. Bradley McWilliams School of Biomedical Informatics, The University of Texas Health Science Center at Houston, Houston, TX, US"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sheikh Muhammad Saiful","family":"Islam","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics and Data Science, Yale School of Medicine, New Haven, CT, US"},{"name":"D. Bradley McWilliams School of Biomedical Informatics, The University of Texas Health Science Center at Houston, Houston, TX, US"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2747-2193","authenticated-orcid":false,"given":"Ardalan","family":"Naseri","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics and Data Science, Yale School of Medicine, New Haven, CT, US"},{"name":"D. Bradley McWilliams School of Biomedical Informatics, The University of Texas Health Science Center at Houston, Houston, TX, US"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhiwen","family":"Fan","sequence":"additional","affiliation":[{"name":"Department of Electrical and Computer Engineering, Texas A&M University, College Station, TX, US"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Han","family":"Chen","sequence":"additional","affiliation":[{"name":"Rory Meyers College of Nursing, NYU, New York, NY, US"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Degui","family":"Zhi","sequence":"additional","affiliation":[{"name":"Department of Biomedical Informatics and Data Science, Yale School of Medicine, New Haven, CT, US"},{"name":"D. Bradley McWilliams School of Biomedical Informatics, The University of Texas Health Science Center at Houston, Houston, TX, US"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"issue":"4","key":"2026100120000783000_2026.09.24.753959v1.1","doi-asserted-by":"crossref","first-page":"749","DOI":"10.1038\/s41592-026-03029-6","article-title":"CellVoyager: AI CompBio agent generates new insights by autonomously analyzing biological data","volume":"23","year":"2026","journal-title":"Nat Methods"},{"issue":"8122","key":"2026100120000783000_2026.09.24.753959v1.2","doi-asserted-by":"crossref","first-page":"497","DOI":"10.1038\/s41586-026-10652-y","article-title":"A multi-agent system for automating scientific discovery","volume":"655","year":"2026","journal-title":"Nature"},{"key":"2026100120000783000_2026.09.24.753959v1.3","doi-asserted-by":"crossref","unstructured":"Huang, K. , et al., Biomni: A General-Purpose Biomedical AI Agent. bioRxiv, 2025.","DOI":"10.1101\/2025.05.30.656746"},{"key":"2026100120000783000_2026.09.24.753959v1.4","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-017-01261-5"},{"issue":"4","key":"2026100120000783000_2026.09.24.753959v1.5","first-page":"332","article-title":"Assessing the heritability of complex traits in humans: methodological challenges and opportunities","volume":"18","year":"2017","journal-title":"Current genomics"},{"key":"2026100120000783000_2026.09.24.753959v1.6","doi-asserted-by":"crossref","first-page":"1557","DOI":"10.1016\/j.csbj.2020.06.011","article-title":"Statistical methods for SNP heritability estimation and partition: A review","volume":"18","year":"2020","journal-title":"Computational and Structural Biotechnology Journal"},{"key":"2026100120000783000_2026.09.24.753959v1.7","doi-asserted-by":"publisher","DOI":"10.1038\/nrg2322"},{"key":"2026100120000783000_2026.09.24.753959v1.8","unstructured":"Fisher, R. , The correlation between relatives on the supposition of mendelian inheritance. Trans. Roy. Soc. 1918."},{"key":"2026100120000783000_2026.09.24.753959v1.9","doi-asserted-by":"crossref","unstructured":"Lande, R. , Quantitative genetic analysis of multivariate evolution, applied to brain: body size allometry. Evolution, 1979: p. 402\u2013416.","DOI":"10.1111\/j.1558-5646.1979.tb04694.x"},{"key":"2026100120000783000_2026.09.24.753959v1.10","doi-asserted-by":"crossref","unstructured":"Hayes, J. and W.G. Hill , A reparameterization of a genetic selection index to locate its sampling properties. Biometrics, 1980: p. 237\u2013248.","DOI":"10.2307\/2529975"},{"key":"2026100120000783000_2026.09.24.753959v1.11","doi-asserted-by":"publisher","DOI":"10.1111\/j.1420-9101.2008.01573.x"},{"key":"2026100120000783000_2026.09.24.753959v1.12","doi-asserted-by":"publisher","DOI":"10.1038\/ncomms13291"},{"key":"2026100120000783000_2026.09.24.753959v1.13","doi-asserted-by":"publisher","DOI":"10.1038\/s41588-021-00827-w"},{"key":"2026100120000783000_2026.09.24.753959v1.14","doi-asserted-by":"publisher","DOI":"10.1007\/BF01066731"},{"issue":"4","key":"2026100120000783000_2026.09.24.753959v1.15","doi-asserted-by":"crossref","first-page":"e1006711","DOI":"10.1371\/journal.pgen.1006711","article-title":"Phenome-wide heritability analysis of the UK Biobank","volume":"13","year":"2017","journal-title":"PLoS genetics"},{"key":"2026100120000783000_2026.09.24.753959v1.16","doi-asserted-by":"crossref","unstructured":"Bulik-Sullivan, B. , Relationship between LD score and Haseman-Elston regression. BioRxiv, 2015: p. 018283.","DOI":"10.1101\/018283"},{"key":"2026100120000783000_2026.09.24.753959v1.17","doi-asserted-by":"publisher","DOI":"10.1038\/ng.3211"},{"key":"2026100120000783000_2026.09.24.753959v1.18","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1419064111"},{"key":"2026100120000783000_2026.09.24.753959v1.19","doi-asserted-by":"publisher","DOI":"10.1159\/000022854"},{"issue":"1","key":"2026100120000783000_2026.09.24.753959v1.20","doi-asserted-by":"crossref","first-page":"414","DOI":"10.1038\/s42003-024-06096-7","article-title":"Unsupervised deep representation learning enables phenotype discovery for genetic association studies of brain imaging","volume":"7","year":"2024","journal-title":"Commun Biol"},{"key":"2026100120000783000_2026.09.24.753959v1.21","doi-asserted-by":"crossref","unstructured":"Islam, S.R. , et al., Vision transformer autoencoders captures local and non-local features in brain imaging to reveal novel genetic associations. Commun Biol, 2026.","DOI":"10.1038\/s42003-026-10430-6"},{"key":"2026100120000783000_2026.09.24.753959v1.22","doi-asserted-by":"publisher","DOI":"10.1038\/s41588-021-00954-4"},{"key":"2026100120000783000_2026.09.24.753959v1.23","doi-asserted-by":"crossref","unstructured":"Guo, B. , et al., Efficient multi-phenotype genome-wide analysis identifies genetic associations for unsupervised deep-learning-derived high-dimensional brain imaging phenotypes. medRxiv, 2024.","DOI":"10.1101\/2024.12.06.24318618"},{"key":"2026100120000783000_2026.09.24.753959v1.24","doi-asserted-by":"crossref","unstructured":"Zhao, X. , et al., Unveiling genetic architecture of white matter microstructure through unsupervised deep representation learning of fractional anisotropy maps. Research Square, 2025: p. rs. 3. rs-7411165.","DOI":"10.1101\/2025.07.04.25330856"},{"issue":"1","key":"2026100120000783000_2026.09.24.753959v1.25","doi-asserted-by":"crossref","first-page":"20","DOI":"10.1002\/gepi.21937","article-title":"USAT: A Unified Score-Based Association Test for Multiple Phenotype-Genotype Analysis","volume":"40","year":"2016","journal-title":"Genet Epidemiol"},{"key":"2026100120000783000_2026.09.24.753959v1.26","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btn563"},{"key":"2026100120000783000_2026.09.24.753959v1.27","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-020-17368-1"},{"key":"2026100120000783000_2026.09.24.753959v1.28","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-018-0579-z"},{"key":"2026100120000783000_2026.09.24.753959v1.29","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-018-0571-7"},{"key":"2026100120000783000_2026.09.24.753959v1.30","doi-asserted-by":"publisher","DOI":"10.1038\/s41593-021-00826-4"},{"key":"2026100120000783000_2026.09.24.753959v1.31","doi-asserted-by":"publisher","DOI":"10.1016\/j.nic.2005.09.008"},{"key":"2026100120000783000_2026.09.24.753959v1.32","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2024.108845"},{"issue":"365","key":"2026100120000783000_2026.09.24.753959v1.33","first-page":"1","article-title":"Advanced normalization tools (ANTS)","volume":"2","year":"2009","journal-title":"Insight j"},{"key":"2026100120000783000_2026.09.24.753959v1.34","doi-asserted-by":"crossref","unstructured":"Islam, S. , et al. Vision Transformer Autoencoders for Unsupervised Representation Learning: Revealing Novel Genetic Associations through Learned Sparse Attention Patterns. in NeurIPS 2025 Workshop for Imageomics: Discovering Biological Knowledge from Images Using AI. 2025.","DOI":"10.1101\/2025.03.24.25324549"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.09.24.753959","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,2]],"date-time":"2026-10-02T03:00:19Z","timestamp":1790910019000},"score":1,"resource":{"primary":{"URL":"http:\/\/biorxiv.org\/lookup\/doi\/10.64898\/2026.09.24.753959"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,29]]},"references-count":34,"URL":"https:\/\/doi.org\/10.64898\/2026.09.24.753959","relation":{},"subject":[],"published":{"date-parts":[[2026,9,29]]},"subtype":"preprint"}}