{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"bioRxiv"}],"indexed":{"date-parts":[[2026,3,28]],"date-time":"2026-03-28T20:48:43Z","timestamp":1774730923105,"version":"3.50.1"},"posted":{"date-parts":[[2026,3,23]]},"group-title":"Systems Biology","reference-count":28,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,3,23]],"date-time":"2026-03-23T00:00:00Z","timestamp":1774224000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100000051","name":"National Human Genome Research Institute","doi-asserted-by":"crossref","award":["U01 HG0120339"],"award-info":[{"award-number":["U01 HG0120339"]}],"id":[{"id":"10.13039\/100000051","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,3,25]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:sec>\n                  <jats:title>Motivation<\/jats:title>\n                  <jats:p>Large-scale gene knockdown\/knockout screens have been used to gain insight into a wide array of phenotypes and biological processes. However, conducting such experiments is expensive and labor-intensive. In this work, we present a general graph-based machine-learning approach that can predict the effects of gene perturbations on molecular phenotypes of interest given some measured phenotypic effects of other gene perturbations. The motivation for learning models that can predict the effects of gene perturbations is fourfold. Such models can (1) predict effects for unmeasured genes in cases in which cost or technical barriers preclude perturbing every gene, (2) prioritize unmeasured genes or sets of genes for subsequent perturbation experiments, (3) hypothesize mechanisms that underlie the relationships between the perturbed genes and their effects, and (4) generalize to other unmeasured phenotypes of interest.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>We evaluate our approach by applying it, in conjunction with four different learning methods, to learn models for four varied phenotypes. Our empirical evaluation demonstrates that the learned models (1) show relatively high levels of predictive accuracy across the four phenotypes, (2) have better predictive accuracy than several standard baselines, (3) can often learn accurate models with small training sets, (4) benefit from having multiple sources of evidence in the input representation, (5) can, in many cases, transfer their predictive value to other phenotypes.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Data availability<\/jats:title>\n                  <jats:p>\n                    The assembled data sets and source code for this work are available at:\n                    <jats:ext-link xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" ext-link-type=\"uri\" xlink:href=\"https:\/\/github.com\/Craven-Biostat-Lab\/graph-molecular-phenotype-prediction\">https:\/\/github.com\/Craven-Biostat-Lab\/graph-molecular-phenotype-prediction<\/jats:ext-link>\n                  <\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Author summary<\/jats:title>\n                  <jats:p>One general approach for gaining insight into the genes involved in a specific biological process is to conduct an experiment in which individual genes are perturbed and the effect on the process is measured for each perturbation. Large-scale experiments of this type have provided important biological insights, but they are often expensive and labor-intensive to perform. As a result, it is not always feasible to measure the effects of perturbing every gene. In this article, we present a machine-learning approach to predicting the effects of gene perturbations using available experimental data and biological network information. Our method can estimate the effects of genes that have not yet been experimentally measured, helping researchers identify promising genes to study next. In addition, the models can suggest hypotheses about the molecular interactions that link genes to the biological process of interest. Approaches like this may help guide experimental studies and accelerate the discovery of gene\u2013phenotype relationships.<\/jats:p>\n                <\/jats:sec>","DOI":"10.64898\/2026.03.20.712202","type":"posted-content","created":{"date-parts":[[2026,3,24]],"date-time":"2026-03-24T04:10:19Z","timestamp":1774325419000},"source":"Crossref","is-referenced-by-count":0,"title":["A graph-based learning approach to predict the effects of gene perturbations on molecular phenotypes"],"prefix":"10.64898","author":[{"given":"Yiyang","family":"Jin","sequence":"first","affiliation":[{"name":"Department of Biostatistics & Medical Informatics, University of Wisconsin, Madison, Wisconsin, U.S.A."}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yuriy","family":"Sverchkov","sequence":"additional","affiliation":[{"name":"Department of Biostatistics & Medical Informatics, University of Wisconsin, Madison, Wisconsin, U.S.A."}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Anastasiya","family":"Sushkova","sequence":"additional","affiliation":[{"name":"Institute for Immunity, Transplantation and Infection, Stanford University, Stanford, California, U.S.A."}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Michael","family":"Ohtake","sequence":"additional","affiliation":[{"name":"Epic Systems, Verona, Wisconsin, U.S.A."}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Christopher","family":"Emfinger","sequence":"additional","affiliation":[{"name":"Barnstable Brown Diabetes Center, University of Kentucky, Lexington, Kentucky, U.S.A."}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Mark","family":"Craven","sequence":"additional","affiliation":[{"name":"Department of Biostatistics & Medical Informatics, University of Wisconsin, Madison, Wisconsin, U.S.A."}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"54368","reference":[{"key":"2026032813000595000_2026.03.20.712202v2.1","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-022-01667-0"},{"key":"2026032813000595000_2026.03.20.712202v2.2","doi-asserted-by":"publisher","DOI":"10.3389\/fcell.2023.1175849"},{"key":"2026032813000595000_2026.03.20.712202v2.3","doi-asserted-by":"publisher","DOI":"10.3389\/fgeed.2023.1243731"},{"key":"2026032813000595000_2026.03.20.712202v2.4","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-022-04765-3"},{"key":"2026032813000595000_2026.03.20.712202v2.5","doi-asserted-by":"publisher","DOI":"10.1186\/s12859-019-2669-9"},{"issue":"9","key":"2026032813000595000_2026.03.20.712202v2.6","doi-asserted-by":"crossref","first-page":"e1008229","DOI":"10.1371\/journal.pcbi.1008229","article-title":"DeepHE: Accurately predicting human essential genes based on deep learning","volume":"16","year":"2020","journal-title":"PLOS Computational Biology"},{"issue":"3","key":"2026032813000595000_2026.03.20.712202v2.7","first-page":"1615","article-title":"EPGAT: gene essentiality prediction with graph attention networks","volume":"19","year":"2021","journal-title":"IEEE\/ACM Transactions on Computational Biology and Bioinformatics"},{"key":"2026032813000595000_2026.03.20.712202v2.8","doi-asserted-by":"crossref","unstructured":"Grover A , Leskovec J. node2vec: Scalable feature learning for networks. In: Proceedings of the 22nd ACM SIGKDD international conference on Knowledge discovery and data mining; 2016. p. 855\u2013864.","DOI":"10.1145\/2939672.2939754"},{"key":"2026032813000595000_2026.03.20.712202v2.9","doi-asserted-by":"publisher","DOI":"10.1038\/s41587-023-01905-6"},{"issue":"1","key":"2026032813000595000_2026.03.20.712202v2.10","first-page":"bbae617","article-title":"BioDSNN: a dual-stream neural network with hybrid biological knowledge integration for multi-gene perturbation response prediction","volume":"26","year":"2025","journal-title":"Briefings in Bioinformatics"},{"key":"2026032813000595000_2026.03.20.712202v2.11","doi-asserted-by":"publisher","DOI":"10.1093\/NAR\/GKAD1025"},{"key":"2026032813000595000_2026.03.20.712202v2.12","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkaa1074"},{"key":"2026032813000595000_2026.03.20.712202v2.13","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkae1010"},{"key":"2026032813000595000_2026.03.20.712202v2.14","doi-asserted-by":"publisher","DOI":"10.1126\/science.1260419"},{"key":"2026032813000595000_2026.03.20.712202v2.15","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/gkh036"},{"key":"2026032813000595000_2026.03.20.712202v2.16","year":"2019","journal-title":"Fast graph representation learning with PyTorch Geometric"},{"issue":"1","key":"2026032813000595000_2026.03.20.712202v2.17","doi-asserted-by":"crossref","first-page":"5559","DOI":"10.1038\/s41467-020-19032-0","article-title":"A trimeric Rab7 GEF controls NPC1-dependent lysosomal cholesterol export","volume":"11","year":"2020","journal-title":"Nature communications"},{"issue":"4","key":"2026032813000595000_2026.03.20.712202v2.18","doi-asserted-by":"crossref","first-page":"279","DOI":"10.1007\/s13238-020-00753-3","article-title":"POST1\/C12ORF49 regulates the SREBP pathway by promoting site-1 protease maturation","volume":"12","year":"2021","journal-title":"Protein & cell"},{"key":"2026032813000595000_2026.03.20.712202v2.19","doi-asserted-by":"crossref","unstructured":"Hamilton MC , Fife JD , Akinci E , Yu T , Khowpinitchai B , Cha M , et al. Systematic elucidation of genetic mechanisms underlying cholesterol uptake. Cell Genomics. 2023;3(5).","DOI":"10.1016\/j.xgen.2023.100304"},{"issue":"2","key":"2026032813000595000_2026.03.20.712202v2.20","doi-asserted-by":"crossref","first-page":"596","DOI":"10.1016\/j.celrep.2018.03.045","article-title":"Genome-wide CRISPR\/Cas9 screen identifies host factors essential for influenza virus replication","volume":"23","year":"2018","journal-title":"Cell reports"},{"issue":"1","key":"2026032813000595000_2026.03.20.712202v2.21","doi-asserted-by":"crossref","first-page":"164","DOI":"10.1038\/s41467-019-13965-x","article-title":"Genome-wide CRISPR screen identifies host dependency factors for influenza A virus infection","volume":"11","year":"2020","journal-title":"Nature communications"},{"key":"2026032813000595000_2026.03.20.712202v2.22","doi-asserted-by":"crossref","unstructured":"Yi C , Cai C , Cheng Z , Zhao Y , Yang X , Wu Y , et al. Genome-wide CRISPR-Cas9 screening identifies the CYTH2 host gene as a potential therapeutic target of influenza viral infection. Cell reports. 2022;38(13).","DOI":"10.1016\/j.celrep.2022.110559"},{"key":"2026032813000595000_2026.03.20.712202v2.23","doi-asserted-by":"publisher","DOI":"10.1186\/gb-2014-15-1-r1"},{"issue":"9","key":"2026032813000595000_2026.03.20.712202v2.24","doi-asserted-by":"crossref","first-page":"e1003235","DOI":"10.1371\/journal.pcbi.1003235","article-title":"Limited Agreement of Indepdendent RNAi Screens for Virus-Required Host Genes Owes More to False-Negative than False-Positive Factors","volume":"9","year":"2013","journal-title":"PLoS Computational Biology"},{"key":"2026032813000595000_2026.03.20.712202v2.25","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467-9868.2005.00503.x"},{"key":"2026032813000595000_2026.03.20.712202v2.26","doi-asserted-by":"publisher","DOI":"10.1023\/A:1010933404324"},{"key":"2026032813000595000_2026.03.20.712202v2.27","doi-asserted-by":"crossref","unstructured":"Chen T , Guestrin C. XGBoost: A scalable tree boosting system. In: Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining; 2016. p. 785\u2013794.","DOI":"10.1145\/2939672.2939785"},{"key":"2026032813000595000_2026.03.20.712202v2.28","doi-asserted-by":"publisher","DOI":"10.1038\/nrg.2017.38"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.03.20.712202","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,28]],"date-time":"2026-03-28T20:00:16Z","timestamp":1774728016000},"score":1,"resource":{"primary":{"URL":"http:\/\/biorxiv.org\/lookup\/doi\/10.64898\/2026.03.20.712202"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,23]]},"references-count":28,"URL":"https:\/\/doi.org\/10.64898\/2026.03.20.712202","relation":{},"subject":[],"published":{"date-parts":[[2026,3,23]]},"subtype":"preprint"}}