{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,4]],"date-time":"2026-06-04T18:13:34Z","timestamp":1780596814635,"version":"3.54.1"},"update-to":[{"DOI":"10.1371\/journal.pcbi.1014327","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2026,6,4]],"date-time":"2026-06-04T00:00:00Z","timestamp":1780531200000}}],"reference-count":52,"publisher":"Public Library of Science (PLoS)","issue":"6","license":[{"start":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T00:00:00Z","timestamp":1780272000000},"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":"publisher","award":["HG013177"],"award-info":[{"award-number":["HG013177"]}],"id":[{"id":"10.13039\/100000051","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000057","name":"National Institute of General Medical Sciences","doi-asserted-by":"publisher","award":["GM105785"],"award-info":[{"award-number":["GM105785"]}],"id":[{"id":"10.13039\/100000057","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["www.ploscompbiol.org"],"crossmark-restriction":false},"short-container-title":["PLoS Comput Biol"],"abstract":"<jats:p>Gene-by-gene differential expression analysis is a widely used supervised approach for interpreting single-cell RNA-sequencing (scRNA-seq) data. However, modern scRNA-seq datasets often contain large numbers of cells, leading to the identification of many differentially expressed genes with extremely small p-values but negligible effect sizes, thus making biological interpretation difficult. To overcome this challenge, we developed Supervised Deep learning with gene functional ANnotation (SDAN), a method that integrates gene functional annotation information (e.g., protein-protein interaction) with gene-expression profiles through a graph neural network. SDAN identifies functionally coherent gene sets that optimally classify cells, and the resulting cell-level classification scores can be aggregated to make individual-level predictions. We evaluated SDAN alongside three representative existing methods in three real-data applications aimed at identifying gene sets associated with severe COVID-19, dementia, and cancer immunotherapy response. Across all applications, SDAN consistently outperformed the alternative approaches by achieving two objectives simultaneously: accurate outcome classification and clear assignment of genes to functionally related gene sets.<\/jats:p>","DOI":"10.1371\/journal.pcbi.1014327","type":"journal-article","created":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T17:56:22Z","timestamp":1780336582000},"page":"e1014327","update-policy":"https:\/\/doi.org\/10.1371\/journal.pcbi.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Supervised deep learning with gene functional annotation for cell classification"],"prefix":"10.1371","volume":"22","author":[{"given":"Zhexiao","family":"Lin","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yuanyuan","family":"Gao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6350-1107","authenticated-orcid":true,"given":"Wei","family":"Sun","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,6,1]]},"reference":[{"issue":"1","key":"pcbi.1014327.ref001","doi-asserted-by":"crossref","first-page":"390","DOI":"10.1038\/s41467-018-07931-2","article-title":"Single-cell RNA-seq denoising using a deep count autoencoder","volume":"10","author":"G Eraslan","year":"2019","journal-title":"Nat Commun"},{"issue":"9","key":"pcbi.1014327.ref002","doi-asserted-by":"crossref","first-page":"875","DOI":"10.1038\/s41592-019-0537-1","article-title":"Data denoising with transfer learning in single-cell transcriptomics","volume":"16","author":"J Wang","year":"2019","journal-title":"Nat Methods"},{"issue":"1","key":"pcbi.1014327.ref003","doi-asserted-by":"crossref","first-page":"112","DOI":"10.1214\/19-STS7560","article-title":"Data denoising and post-denoising corrections in single cell RNA sequencing","volume":"35","author":"D Agarwal","year":"2020","journal-title":"Statistical Science"},{"issue":"4","key":"pcbi.1014327.ref004","doi-asserted-by":"crossref","first-page":"191","DOI":"10.1038\/s42256-019-0037-0","article-title":"Clustering single-cell RNA-seq data with a model-based deep learning approach","volume":"1","author":"T Tian","year":"2019","journal-title":"Nat Mach Intell"},{"issue":"1","key":"pcbi.1014327.ref005","doi-asserted-by":"crossref","first-page":"2338","DOI":"10.1038\/s41467-020-15851-3","article-title":"Deep learning enables accurate clustering with batch effect removal in single-cell RNA-seq analysis","volume":"11","author":"X Li","year":"2020","journal-title":"Nat Commun"},{"issue":"12","key":"pcbi.1014327.ref006","doi-asserted-by":"crossref","first-page":"1053","DOI":"10.1038\/s41592-018-0229-2","article-title":"Deep generative modeling for single-cell transcriptomics","volume":"15","author":"R Lopez","year":"2018","journal-title":"Nat Methods"},{"issue":"1","key":"pcbi.1014327.ref007","article-title":"Probabilistic harmonization and annotation of single-cell transcriptomics data with deep generative models","volume":"17","author":"C Xu","year":"2021","journal-title":"Mol Syst Biol"},{"key":"pcbi.1014327.ref008","doi-asserted-by":"crossref","DOI":"10.7554\/eLife.43803","article-title":"Identifying gene expression programs of cell-type identity and cellular activity with single-cell RNA-Seq","volume":"8","author":"D Kotliar","year":"2019","journal-title":"Elife"},{"issue":"1","key":"pcbi.1014327.ref009","doi-asserted-by":"crossref","first-page":"1878","DOI":"10.1038\/s41467-025-57157-2","article-title":"Interpretable single-cell factor decomposition using sciRED","volume":"16","author":"D Pouyabahar","year":"2025","journal-title":"Nat Commun"},{"issue":"7","key":"pcbi.1014327.ref010","doi-asserted-by":"crossref","first-page":"1084","DOI":"10.1038\/s41587-023-01940-3","article-title":"Supervised discovery of interpretable gene programs from single-cell data","volume":"42","author":"RZ Kunes","year":"2024","journal-title":"Nat Biotechnol"},{"issue":"4","key":"pcbi.1014327.ref011","doi-asserted-by":"crossref","first-page":"708","DOI":"10.1038\/s41592-025-02627-0","article-title":"scNET: learning context-specific gene and cell embeddings by integrating single-cell gene expression data with protein-protein interactions","volume":"22","author":"R Sheinin","year":"2025","journal-title":"Nat Methods"},{"issue":"11","key":"pcbi.1014327.ref012","doi-asserted-by":"crossref","first-page":"2032","DOI":"10.1038\/s41591-021-01540-1","article-title":"Correlates of protection against symptomatic and asymptomatic SARS-CoV-2 infection","volume":"27","author":"S Feng","year":"2021","journal-title":"Nat Med"},{"issue":"7","key":"pcbi.1014327.ref013","doi-asserted-by":"crossref","first-page":"1205","DOI":"10.1038\/s41591-021-01377-8","article-title":"Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection","volume":"27","author":"DS Khoury","year":"2021","journal-title":"Nat Med"},{"issue":"6576","key":"pcbi.1014327.ref014","doi-asserted-by":"crossref","first-page":"43","DOI":"10.1126\/science.abm3425","article-title":"Immune correlates analysis of the mRNA-1273 COVID-19 vaccine efficacy clinical trial","volume":"375","author":"PB Gilbert","year":"2022","journal-title":"Science"},{"issue":"11","key":"pcbi.1014327.ref015","doi-asserted-by":"crossref","first-page":"1085","DOI":"10.1056\/NEJMc2025179","article-title":"Rapid Decay of Anti-SARS-CoV-2 Antibodies in Persons with Mild Covid-19","volume":"383","author":"FJ Ibarrondo","year":"2020","journal-title":"N Engl J Med"},{"issue":"3","key":"pcbi.1014327.ref016","doi-asserted-by":"crossref","DOI":"10.1016\/j.cell.2021.12.032","article-title":"The Omicron variant is highly resistant against antibody-mediated neutralization: Implications for control of the COVID-19 pandemic","volume":"185","author":"M Hoffmann","year":"2022","journal-title":"Cell"},{"issue":"4","key":"pcbi.1014327.ref017","doi-asserted-by":"crossref","first-page":"861","DOI":"10.1016\/j.cell.2021.01.007","article-title":"Adaptive immunity to SARS-CoV-2 and COVID-19","volume":"184","author":"A Sette","year":"2021","journal-title":"Cell"},{"issue":"4","key":"pcbi.1014327.ref018","doi-asserted-by":"crossref","DOI":"10.1016\/j.cell.2020.09.038","article-title":"Antigen-Specific Adaptive Immunity to SARS-CoV-2 in Acute COVID-19 and Associations with Age and Disease Severity","volume":"183","author":"C Rydyznski Moderbacher","year":"2020","journal-title":"Cell"},{"issue":"6","key":"pcbi.1014327.ref019","doi-asserted-by":"crossref","first-page":"108728","DOI":"10.1016\/j.celrep.2021.108728","article-title":"Early induction of functional SARS-CoV-2-specific T cells associates with rapid viral clearance and mild disease in COVID-19 patients","volume":"34","author":"AT Tan","year":"2021","journal-title":"Cell Rep"},{"issue":"7","key":"pcbi.1014327.ref020","doi-asserted-by":"crossref","first-page":"1280","DOI":"10.1038\/s41591-021-01386-7","article-title":"CD8+ T cells contribute to survival in patients with COVID-19 and hematologic cancer","volume":"27","author":"EM Bange","year":"2021","journal-title":"Nat Med"},{"issue":"7875","key":"pcbi.1014327.ref021","doi-asserted-by":"crossref","first-page":"268","DOI":"10.1038\/s41586-021-03841-4","article-title":"Rapid and stable mobilization of CD8+ T cells by SARS-CoV-2 mRNA vaccine","volume":"597","author":"V Oberhardt","year":"2021","journal-title":"Nature"},{"issue":"5","key":"pcbi.1014327.ref022","doi-asserted-by":"crossref","DOI":"10.1016\/j.cell.2022.01.015","article-title":"SARS-CoV-2 vaccination induces immunological T cell memory able to cross-recognize variants from Alpha to Omicron","volume":"185","author":"A Tarke","year":"2022","journal-title":"Cell"},{"issue":"2","key":"pcbi.1014327.ref023","doi-asserted-by":"crossref","first-page":"186","DOI":"10.1038\/s41590-021-01122-w","article-title":"The T cell immune response against SARS-CoV-2","volume":"23","author":"P Moss","year":"2022","journal-title":"Nat Immunol"},{"issue":"6","key":"pcbi.1014327.ref024","doi-asserted-by":"crossref","DOI":"10.1016\/j.cell.2020.08.001","article-title":"Severe COVID-19 Is Marked by a Dysregulated Myeloid Cell Compartment","volume":"182","author":"J Schulte-Schrepping","year":"2020","journal-title":"Cell"},{"issue":"6","key":"pcbi.1014327.ref025","article-title":"Multi-Omics Resolves a Sharp Disease-State Shift between Mild and Moderate COVID-19","volume":"183","author":"Y Su","year":"2020","journal-title":"Cell"},{"issue":"4","key":"pcbi.1014327.ref026","doi-asserted-by":"crossref","DOI":"10.1016\/j.devcel.2020.12.010","article-title":"ORF3a of the COVID-19 virus SARS-CoV-2 blocks HOPS complex-mediated assembly of the SNARE complex required for autolysosome formation","volume":"56","author":"G Miao","year":"2021","journal-title":"Dev Cell"},{"key":"pcbi.1014327.ref027","doi-asserted-by":"crossref","first-page":"15056","DOI":"10.1038\/nrdp.2015.56","article-title":"Alzheimer\u2019s disease","volume":"1","author":"CL Masters","year":"2015","journal-title":"Nat Rev Dis Primers"},{"issue":"5","key":"pcbi.1014327.ref028","doi-asserted-by":"crossref","first-page":"321","DOI":"10.1038\/s41577-024-01104-7","article-title":"Neuroinflammation in Alzheimer disease","volume":"25","author":"MT Heneka","year":"2025","journal-title":"Nat Rev Immunol"},{"issue":"6286","key":"pcbi.1014327.ref029","doi-asserted-by":"crossref","first-page":"712","DOI":"10.1126\/science.aad8373","article-title":"Complement and microglia mediate early synapse loss in Alzheimer mouse models","volume":"352","author":"S Hong","year":"2016","journal-title":"Science"},{"issue":"6","key":"pcbi.1014327.ref030","doi-asserted-by":"crossref","first-page":"701","DOI":"10.1038\/s41593-020-0624-8","article-title":"Disease-associated astrocytes in Alzheimer\u2019s disease and aging","volume":"23","author":"N Habib","year":"2020","journal-title":"Nat Neurosci"},{"issue":"7638","key":"pcbi.1014327.ref031","doi-asserted-by":"crossref","first-page":"481","DOI":"10.1038\/nature21029","article-title":"Neurotoxic reactive astrocytes are induced by activated microglia","volume":"541","author":"SA Liddelow","year":"2017","journal-title":"Nature"},{"key":"pcbi.1014327.ref032","doi-asserted-by":"crossref","first-page":"195","DOI":"10.1016\/0304-3940(87)90696-3","article-title":"Reactive microglia in patients with senile dementia of the Alzheimer type are positive for the histocompatibility glycoprotein HLA-DR","volume":"79","author":"PL McGeer","year":"1987","journal-title":"Neurosci Lett"},{"issue":"9","key":"pcbi.1014327.ref033","doi-asserted-by":"crossref","first-page":"1276","DOI":"10.1038\/s41588-021-00921-z","article-title":"A genome-wide association study with 1,126,563 individuals identifies new risk loci for Alzheimer\u2019s disease","volume":"53","author":"DP Wightman","year":"2021","journal-title":"Nat Genet"},{"issue":"12","key":"pcbi.1014327.ref034","doi-asserted-by":"crossref","first-page":"2366","DOI":"10.1038\/s41593-024-01774-5","article-title":"Integrated multimodal cell atlas of Alzheimer\u2019s disease","volume":"27","author":"MI Gabitto","year":"2024","journal-title":"Nat Neurosci"},{"issue":"3","key":"pcbi.1014327.ref035","doi-asserted-by":"crossref","first-page":"133","DOI":"10.1038\/s41568-019-0116-x","article-title":"The evolving landscape of biomarkers for checkpoint inhibitor immunotherapy","volume":"19","author":"JJ Havel","year":"2019","journal-title":"Nat Rev Cancer"},{"issue":"4","key":"pcbi.1014327.ref036","doi-asserted-by":"crossref","DOI":"10.1016\/j.cell.2018.10.038","article-title":"Defining T Cell States Associated with Response to Checkpoint Immunotherapy in Melanoma","volume":"175","author":"M Sade-Feldman","year":"2018","journal-title":"Cell"},{"issue":"8","key":"pcbi.1014327.ref037","doi-asserted-by":"crossref","first-page":"1251","DOI":"10.1038\/s41591-019-0522-3","article-title":"Clonal replacement of tumor-specific T cells following PD-1 blockade","volume":"25","author":"KE Yost","year":"2019","journal-title":"Nat Med"},{"issue":"5","key":"pcbi.1014327.ref038","doi-asserted-by":"crossref","first-page":"488","DOI":"10.1038\/ni.3704","article-title":"Mitochondria are the powerhouses of immunity","volume":"18","author":"EL Mills","year":"2017","journal-title":"Nat Immunol"},{"key":"pcbi.1014327.ref039","doi-asserted-by":"crossref","DOI":"10.7554\/eLife.52330","article-title":"Tumors attenuating the mitochondrial activity in T cells escape from PD-1 blockade therapy","volume":"9","author":"A Kumar","year":"2020","journal-title":"Elife"},{"issue":"1","key":"pcbi.1014327.ref040","doi-asserted-by":"crossref","first-page":"6858","DOI":"10.1038\/s41467-023-42634-3","article-title":"Mitochondrial dysfunction promotes the transition of precursor to terminally exhausted T cells through HIF-1\u03b1-mediated glycolytic reprogramming","volume":"14","author":"H Wu","year":"2023","journal-title":"Nat Commun"},{"issue":"5","key":"pcbi.1014327.ref041","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.1620433114","article-title":"Mitochondrial activation chemicals synergize with surface receptor PD-1 blockade for T cell-dependent antitumor activity","volume":"114","author":"K Chamoto","year":"2017","journal-title":"Proc Natl Acad Sci U S A"},{"key":"pcbi.1014327.ref042","doi-asserted-by":"crossref","first-page":"151","DOI":"10.1016\/j.mito.2021.12.001","article-title":"Mitochondria determine response to anti-programmed cell death protein-1 (anti-PD-1) immunotherapy: An evidence-based hypothesis","volume":"62","author":"H Akbari","year":"2022","journal-title":"Mitochondrion"},{"issue":"4","key":"pcbi.1014327.ref043","doi-asserted-by":"crossref","first-page":"290","DOI":"10.1038\/nmeth.4627","article-title":"Using deep learning to model the hierarchical structure and function of a cell","volume":"15","author":"J Ma","year":"2018","journal-title":"Nat Methods"},{"issue":"1","key":"pcbi.1014327.ref044","doi-asserted-by":"crossref","first-page":"510","DOI":"10.1186\/s12859-018-2500-z","article-title":"PASNet: pathway-associated sparse deep neural network for prognosis prediction from high-throughput data","volume":"19","author":"J Hao","year":"2018","journal-title":"BMC Bioinformatics"},{"issue":"7880","key":"pcbi.1014327.ref045","doi-asserted-by":"crossref","first-page":"348","DOI":"10.1038\/s41586-021-03922-4","article-title":"Biologically informed deep neural network for prostate cancer discovery","volume":"598","author":"HA Elmarakeby","year":"2021","journal-title":"Nature"},{"key":"pcbi.1014327.ref046","doi-asserted-by":"crossref","unstructured":"Lotfollahi M, Rybakov S, Hrovatin K, Hediyeh-zadeh S, Talavera-L\u00f3pez C, Misharin A. Biologically informed deep learning to infer gene program activity in single cells. bioRxiv. 2022.","DOI":"10.1101\/2022.02.05.479217"},{"key":"pcbi.1014327.ref047","article-title":"Co-teaching: Robust training of deep neural networks with extremely noisy labels","volume":"31","author":"B Han","year":"2018","journal-title":"Advances in Neural Information Processing Systems"},{"key":"pcbi.1014327.ref048","unstructured":"Yu X, Han B, Yao J, Niu G, Tsang I, Sugiyama M. In: International Conference on Machine Learning, 2019. 7164\u201373."},{"key":"pcbi.1014327.ref049","unstructured":"Li J, Socher R, Hoi SC. DivideMix: Learning with Noisy Labels as Semi-supervised Learning. In: 2019."},{"issue":"7","key":"pcbi.1014327.ref050","doi-asserted-by":"crossref","first-page":"976","DOI":"10.1093\/bioinformatics\/btq064","article-title":"GOSemSim: an R package for measuring semantic similarity among GO terms and gene products","volume":"26","author":"G Yu","year":"2010","journal-title":"Bioinformatics"},{"key":"pcbi.1014327.ref051","unstructured":"Kipf TN, Welling M. Semi-supervised classification with graph convolutional networks. 2016. https:\/\/arxiv.org\/abs\/1609.02907"},{"key":"pcbi.1014327.ref052","unstructured":"Bianchi FM, Grattarola D, Alippi C. In: 2020. 874\u201383."}],"updated-by":[{"DOI":"10.1371\/journal.pcbi.1014327","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2026,6,4]],"date-time":"2026-06-04T00:00:00Z","timestamp":1780531200000}}],"container-title":["PLOS Computational Biology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1014327","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,4]],"date-time":"2026-06-04T17:44:05Z","timestamp":1780595045000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1014327"}},"subtitle":[],"editor":[{"given":"Peng","family":"Wei","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,6,1]]},"references-count":52,"journal-issue":{"issue":"6","published-online":{"date-parts":[[2026,6,1]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pcbi.1014327","relation":{},"ISSN":["1553-7358"],"issn-type":[{"value":"1553-7358","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,1]]}}}