{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,13]],"date-time":"2026-10-13T18:44:30Z","timestamp":1791917070368,"version":"4.3.4"},"reference-count":47,"publisher":"Public Library of Science (PLoS)","issue":"10","license":[{"start":{"date-parts":[[2026,10,13]],"date-time":"2026-10-13T00:00:00Z","timestamp":1791849600000},"content-version":"am","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"NIH","award":["R01NS139485"],"award-info":[{"award-number":["R01NS139485"]}]},{"name":"NIH","award":["R01CA301708"],"award-info":[{"award-number":["R01CA301708"]}]}],"content-domain":{"domain":["www.ploscompbiol.org"],"crossmark-restriction":false},"short-container-title":["PLoS Comput Biol"],"abstract":"<jats:p>Identifying cancer driver genes and their therapeutic impact remains a core challenge in computational cancer biology. We introduce xNNDriver and xAEDriver, two interpretable neural network frameworks that connect cancer mutations with genome-wide DepMap gene dependencies, pathway activity, and drug-response patterns. xNNDriver is a supervised pathway-guided model that evaluates whether a gene\u2019s mutation status is encoded in the genome-wide dependency landscape; we interpret model fitness as a driver potential score, which quantifies the strength of this mutation-dependency signal and prioritizes genes with broad functional footprints. Across 3,008 candidate genes, xNNDriver recovers major established drivers and highlights literature-supported candidates, while pathway analyses reveal biologically coherent programs related to metabolism, growth factor signaling, and immune regulation. To capture combinatorial functional states, xAEDriver uses an unsupervised autoencoder to learn Driver Variant Representations (DVRs), latent binary features guided by the frequency distribution of known driver mutations. DVRs capture cell-line-specific dependency patterns and expression patterns and are associated with drug sensitivity and pathway activity. Together, these interpretable deep learning models demonstrate that gene dependency landscapes encode rich, interpretable signals of oncogenic function and provide a hypothesis-generating framework for prioritizing drivers, pathways, and therapeutic vulnerabilities for further experimental validation.<\/jats:p>","DOI":"10.1371\/journal.pcbi.1014869","type":"journal-article","created":{"date-parts":[[2026,10,13]],"date-time":"2026-10-13T17:46:01Z","timestamp":1791913561000},"page":"e1014869","update-policy":"https:\/\/doi.org\/10.1371\/journal.pcbi.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Interpretable neural networks prioritize cancer driver genes from genome-wide dependency landscapes"],"prefix":"10.1371","volume":"22","author":[{"ORCID":"https:\/\/orcid.org\/0009-0004-3643-7981","authenticated-orcid":true,"given":"Qingyang","family":"Yin","sequence":"first","affiliation":[{"name":"University of Southern California","place":["Los Angeles, California, United States of America"],"department":["Department of Quantitative and Computational Biology"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6164-4553","authenticated-orcid":true,"given":"Liang","family":"Chen","sequence":"additional","affiliation":[{"name":"University of Southern California","place":["Los Angeles, California, United States of America"],"department":["Department of Quantitative and Computational Biology"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,10,13]]},"reference":[{"issue":"24","key":"pcbi.1014869.ref001","doi-asserted-by":"crossref","first-page":"4233","DOI":"10.1002\/1873-3468.13781","article-title":"Understanding oncogenicity of cancer driver genes and mutations in the cancer genomics era","volume":"594","author":"E Porta-Pardo","year":"2020","journal-title":"FEBS Lett"},{"key":"pcbi.1014869.ref002","doi-asserted-by":"crossref","DOI":"10.1093\/nar\/gkad295","article-title":"DeepAlloDriver: a deep learning-based strategy to predict cancer driver mutations","volume":"51","author":"Q Song","year":"2023","journal-title":"Nucleic Acids Res"},{"key":"pcbi.1014869.ref003","doi-asserted-by":"crossref","first-page":"491","DOI":"10.1186\/s12859-021-04400-4","article-title":"Deep learning for cancer type classification and driver gene identification","volume":"22","author":"Z Zeng","year":"2021","journal-title":"BMC Bioinformatics"},{"issue":"1","key":"pcbi.1014869.ref004","doi-asserted-by":"crossref","DOI":"10.1093\/bib\/bbab432","article-title":"Improving cancer driver gene identification using multi-task learning on graph convolutional network","volume":"23","author":"W Peng","year":"2022","journal-title":"Brief Bioinform"},{"issue":"2","key":"pcbi.1014869.ref005","article-title":"Inference of marker genes of subtle cell state changes via iLR: iterative logistic regression","volume":"42","author":"Y Liu","year":"2026","journal-title":"Bioinformatics"},{"issue":"1","key":"pcbi.1014869.ref006","article-title":"scIAE: an integrative autoencoder-based ensemble classification framework for single-cell RNA-seq data","volume":"23","author":"Q Yin","year":"2022","journal-title":"Brief Bioinform"},{"key":"pcbi.1014869.ref007","doi-asserted-by":"crossref","first-page":"57","DOI":"10.1016\/j.gde.2020.12.005","article-title":"The use of CRISPR\/Cas9-based gene editing strategies to explore cancer gene function in mice","volume":"66","author":"L van der Weyden","year":"2021","journal-title":"Curr Opin Genet Dev"},{"issue":"34","key":"pcbi.1014869.ref008","article-title":"Predicting and characterizing a cancer dependency map of tumors with deep learning","volume":"7","author":"Y-C Chiu","year":"2021","journal-title":"Sci Adv"},{"key":"pcbi.1014869.ref009","article-title":"Extracting biological insights from the project achilles genome-scale CRISPR screens in cancer cell lines","author":"JM Dempster","year":"2019","journal-title":"BioRxiv"},{"issue":"12","key":"pcbi.1014869.ref010","doi-asserted-by":"crossref","first-page":"1779","DOI":"10.1038\/ng.3984","article-title":"Computational correction of copy number effect improves specificity of CRISPR-Cas9 essentiality screens in cancer cells","volume":"49","author":"RM Meyers","year":"2017","journal-title":"Nat Genet"},{"key":"pcbi.1014869.ref011","article-title":"Reactome: a knowledgebase of biological pathways","volume":"33","author":"G Joshi-Tope","year":"2005","journal-title":"Nucleic Acids Res"},{"key":"pcbi.1014869.ref012","doi-asserted-by":"crossref","DOI":"10.1093\/nar\/gkq929","article-title":"COSMIC: mining complete cancer genomes in the Catalogue of Somatic Mutations in Cancer","volume":"39","author":"SA Forbes","year":"2011","journal-title":"Nucleic Acids Res"},{"key":"pcbi.1014869.ref013","doi-asserted-by":"crossref","DOI":"10.1093\/nar\/gkt1113","article-title":"ClinVar: public archive of relationships among sequence variation and human phenotype","volume":"42","author":"MJ Landrum","year":"2014","journal-title":"Nucleic Acids Res"},{"issue":"5","key":"pcbi.1014869.ref014","doi-asserted-by":"crossref","DOI":"10.1016\/j.ccell.2020.09.014","article-title":"Predicting Drug Response and Synergy Using a Deep Learning Model of Human Cancer Cells","volume":"38","author":"BM Kuenzi","year":"2020","journal-title":"Cancer Cell"},{"issue":"3","key":"pcbi.1014869.ref015","doi-asserted-by":"crossref","first-page":"873","DOI":"10.1158\/0008-5472.CAN-10-2164","article-title":"LATS2 is a tumor suppressor gene of malignant mesothelioma","volume":"71","author":"H Murakami","year":"2011","journal-title":"Cancer Res"},{"issue":"3","key":"pcbi.1014869.ref016","doi-asserted-by":"crossref","first-page":"257","DOI":"10.1016\/j.lungcan.2008.09.011","article-title":"LATS2 tumour specific mutations and down-regulation of the gene in non-small cell carcinoma","volume":"64","author":"M Strazisar","year":"2009","journal-title":"Lung Cancer"},{"key":"pcbi.1014869.ref017","doi-asserted-by":"crossref","first-page":"1233953","DOI":"10.3389\/fonc.2023.1233953","article-title":"Non-canonical functions of EZH2 in cancer","volume":"13","author":"SM Zimmerman","year":"2023","journal-title":"Front Oncol"},{"issue":"3","key":"pcbi.1014869.ref018","doi-asserted-by":"crossref","first-page":"719","DOI":"10.3892\/ol.2013.1427","article-title":"Keap1 mutations in lung cancer patients","volume":"6","author":"H Sasaki","year":"2013","journal-title":"Oncol Lett"},{"issue":"12","key":"pcbi.1014869.ref019","doi-asserted-by":"crossref","first-page":"1531","DOI":"10.1530\/EC-21-0410","article-title":"BRAF V600E mutation in papillary thyroid carcinoma: it\u2019s relation to clinical features and oncologic outcomes in a single cancer centre experience","volume":"10","author":"M Al-Masri","year":"2021","journal-title":"Endocr Connect"},{"issue":"12","key":"pcbi.1014869.ref020","doi-asserted-by":"crossref","first-page":"2006","DOI":"10.3390\/cancers11122006","article-title":"Somatic Mutation of PIK3CA (H1047R) Is a Common Driver Mutation Hotspot in Canine Mammary Tumors as Well as Human Breast Cancers","volume":"11","author":"K-H Lee","year":"2019","journal-title":"Cancers (Basel)"},{"issue":"4","key":"pcbi.1014869.ref021","doi-asserted-by":"crossref","first-page":"657","DOI":"10.1016\/j.cell.2016.12.039","article-title":"Understanding the Intersections between Metabolism and Cancer Biology","volume":"168","author":"MG Vander Heiden","year":"2017","journal-title":"Cell"},{"issue":"4","key":"pcbi.1014869.ref022","doi-asserted-by":"crossref","first-page":"264","DOI":"10.1002\/mc.20665","article-title":"Growth factor signaling pathways as targets for prevention of epithelial carcinogenesis","volume":"50","author":"O Rho","year":"2011","journal-title":"Mol Carcinog"},{"issue":"7","key":"pcbi.1014869.ref023","doi-asserted-by":"crossref","first-page":"1117","DOI":"10.1016\/j.cell.2010.06.011","article-title":"Cell signaling by receptor tyrosine kinases","volume":"141","author":"MA Lemmon","year":"2010","journal-title":"Cell"},{"issue":"8","key":"pcbi.1014869.ref024","doi-asserted-by":"crossref","first-page":"553","DOI":"10.1038\/nrc3309","article-title":"The ERBB network: at last, cancer therapy meets systems biology","volume":"12","author":"Y Yarden","year":"2012","journal-title":"Nat Rev Cancer"},{"issue":"6","key":"pcbi.1014869.ref025","doi-asserted-by":"crossref","first-page":"883","DOI":"10.1016\/j.cell.2010.01.025","article-title":"Immunity, inflammation, and cancer","volume":"140","author":"SI Grivennikov","year":"2010","journal-title":"Cell"},{"issue":"7203","key":"pcbi.1014869.ref026","doi-asserted-by":"crossref","first-page":"436","DOI":"10.1038\/nature07205","article-title":"Cancer-related inflammation","volume":"454","author":"A Mantovani","year":"2008","journal-title":"Nature"},{"issue":"9","key":"pcbi.1014869.ref027","doi-asserted-by":"crossref","first-page":"11553","DOI":"10.1007\/s13277-016-5098-7","article-title":"Role of interleukin-6 in cancer progression and therapeutic resistance","volume":"37","author":"N Kumari","year":"2016","journal-title":"Tumour Biol"},{"issue":"1","key":"pcbi.1014869.ref028","article-title":"CellTICS: an explainable neural network for cell-type identification and interpretation based on single-cell RNA-seq data","volume":"25","author":"Q Yin","year":"2023","journal-title":"Brief Bioinform"},{"key":"pcbi.1014869.ref029","doi-asserted-by":"crossref","unstructured":"Rubner Y, Tomasi C, Guibas LJ. A metric for distributions with applications to image databases. In: Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271). 59\u201366. https:\/\/doi.org\/10.1109\/iccv.1998.710701","DOI":"10.1109\/ICCV.1998.710701"},{"issue":"11","key":"pcbi.1014869.ref030","doi-asserted-by":"crossref","first-page":"1453","DOI":"10.1158\/1541-7786.MCR-10-0225","article-title":"Cancer and the complement cascade","volume":"8","author":"MJ Rutkowski","year":"2010","journal-title":"Mol Cancer Res"},{"issue":"1","key":"pcbi.1014869.ref031","doi-asserted-by":"crossref","first-page":"292","DOI":"10.1038\/s41419-025-07609-w","article-title":"Inflammation and cancer cell survival: TRAF2 as a key player","volume":"16","author":"A Albini","year":"2025","journal-title":"Cell Death Dis"},{"key":"pcbi.1014869.ref032","doi-asserted-by":"crossref","first-page":"101049","DOI":"10.1016\/j.neo.2024.101049","article-title":"ELAVL1 regulates PD-L1 mRNA stability to disrupt the infiltration of CD4-positive T cells in prostate cancer","volume":"57","author":"Z Cai","year":"2024","journal-title":"Neoplasia"},{"issue":"17","key":"pcbi.1014869.ref033","doi-asserted-by":"crossref","first-page":"6987","DOI":"10.1158\/0008-5472.CAN-09-1106","article-title":"Dual therapeutic efficacy of vinblastine as a unique chemotherapeutic agent capable of inducing dendritic cell maturation","volume":"69","author":"H Tanaka","year":"2009","journal-title":"Cancer Res"},{"issue":"1","key":"pcbi.1014869.ref034","doi-asserted-by":"crossref","first-page":"272","DOI":"10.1186\/s12864-021-07581-7","article-title":"Predicting tumor response to drugs based on gene-expression biomarkers of sensitivity learned from cancer cell lines","volume":"22","author":"Y Li","year":"2021","journal-title":"BMC Genomics"},{"issue":"9","key":"pcbi.1014869.ref035","doi-asserted-by":"crossref","first-page":"4228","DOI":"10.4049\/jimmunol.1500622","article-title":"Immunosuppression and Aberrant T Cell Development in the Absence of N-Myristoylation","volume":"195","author":"F Rampoldi","year":"2015","journal-title":"J Immunol"},{"issue":"1","key":"pcbi.1014869.ref036","doi-asserted-by":"crossref","first-page":"51","DOI":"10.1038\/s41698-022-00298-0","article-title":"Targeting Mre11 overcomes platinum resistance and induces synthetic lethality in XRCC1 deficient epithelial ovarian cancers","volume":"6","author":"A Alblihy","year":"2022","journal-title":"NPJ Precis Oncol"},{"issue":"12","key":"pcbi.1014869.ref037","doi-asserted-by":"crossref","first-page":"3655","DOI":"10.1007\/s12033-023-00973-y","article-title":"Identification of Crucial Genes and Signaling Pathways in Alectinib-Resistant Lung Adenocarcinoma Using Bioinformatic Analysis","volume":"66","author":"Z Li","year":"2024","journal-title":"Mol Biotechnol"},{"issue":"3","key":"pcbi.1014869.ref038","doi-asserted-by":"crossref","DOI":"10.1016\/j.cell.2017.06.010","article-title":"Defining a Cancer Dependency Map","volume":"170","author":"A Tsherniak","year":"2017","journal-title":"Cell"},{"issue":"7753","key":"pcbi.1014869.ref039","doi-asserted-by":"crossref","first-page":"511","DOI":"10.1038\/s41586-019-1103-9","article-title":"Prioritization of cancer therapeutic targets using CRISPR-Cas9 screens","volume":"568","author":"FM Behan","year":"2019","journal-title":"Nature"},{"issue":"7391","key":"pcbi.1014869.ref040","doi-asserted-by":"crossref","first-page":"603","DOI":"10.1038\/nature11003","article-title":"The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity","volume":"483","author":"J Barretina","year":"2012","journal-title":"Nature"},{"issue":"7757","key":"pcbi.1014869.ref041","doi-asserted-by":"crossref","first-page":"503","DOI":"10.1038\/s41586-019-1186-3","article-title":"Next-generation characterization of the Cancer Cell Line Encyclopedia","volume":"569","author":"M Ghandi","year":"2019","journal-title":"Nature"},{"issue":"3","key":"pcbi.1014869.ref042","doi-asserted-by":"crossref","first-page":"740","DOI":"10.1016\/j.cell.2016.06.017","article-title":"A Landscape of Pharmacogenomic Interactions in Cancer","volume":"166","author":"F Iorio","year":"2016","journal-title":"Cell"},{"issue":"7","key":"pcbi.1014869.ref043","doi-asserted-by":"crossref","first-page":"691","DOI":"10.2217\/pgs.16.15","article-title":"Multilevel models improve precision and speed of IC50 estimates","volume":"17","author":"DJ Vis","year":"2016","journal-title":"Pharmacogenomics"},{"issue":"12","key":"pcbi.1014869.ref044","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pone.0146021","article-title":"Dose-Response Analysis Using R","volume":"10","author":"C Ritz","year":"2015","journal-title":"PLoS One"},{"key":"pcbi.1014869.ref045","doi-asserted-by":"crossref","DOI":"10.1093\/nar\/gkad1049","article-title":"Ensembl 2024","volume":"52","author":"PW Harrison","year":"2024","journal-title":"Nucleic Acids Res"},{"issue":"5","key":"pcbi.1014869.ref046","doi-asserted-by":"crossref","DOI":"10.1186\/gb-2003-4-5-p3","article-title":"DAVID: Database for Annotation, Visualization, and Integrated Discovery","volume":"4","author":"G Dennis","year":"2003","journal-title":"Genome Biol"},{"key":"pcbi.1014869.ref047","doi-asserted-by":"crossref","unstructured":"Binder A, Montavon G, Lapuschkin S, M\u00fcller KR, Samek W. Layer-wise relevance propagation for neural networks with local renormalization layers. In: 2016.","DOI":"10.1007\/978-3-319-44781-0_8"}],"container-title":["PLOS Computational Biology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1014869","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,13]],"date-time":"2026-10-13T17:46:10Z","timestamp":1791913570000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1014869"}},"subtitle":[],"editor":[{"given":"Hang","family":"Wei","sequence":"first","affiliation":[{"name":"Xidian University","place":["CHINA"]}],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,10,13]]},"references-count":47,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2026,10,13]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pcbi.1014869","relation":{},"ISSN":["1553-7358"],"issn-type":[{"value":"1553-7358","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,13]]}}}