{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,27]],"date-time":"2026-04-27T12:37:24Z","timestamp":1777293444233,"version":"3.51.4"},"update-to":[{"DOI":"10.1371\/journal.pcbi.1011476","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2023,10,12]],"date-time":"2023-10-12T00:00:00Z","timestamp":1697068800000}}],"reference-count":55,"publisher":"Public Library of Science (PLoS)","issue":"10","license":[{"start":{"date-parts":[[2023,10,2]],"date-time":"2023-10-02T00:00:00Z","timestamp":1696204800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100000038","name":"NSERC","doi-asserted-by":"crossref","award":["RGPIN-2018-05147"],"award-info":[{"award-number":["RGPIN-2018-05147"]}],"id":[{"id":"10.13039\/501100000038","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/100008459","name":"University of Calgary","doi-asserted-by":"crossref","award":["VPR Catalyst grant"],"award-info":[{"award-number":["VPR Catalyst grant"]}],"id":[{"id":"10.13039\/100008459","id-type":"DOI","asserted-by":"crossref"}]},{"name":"New Frontiers in Research Fund","award":["NFRFE-2018-00748"],"award-info":[{"award-number":["NFRFE-2018-00748"]}]},{"DOI":"10.13039\/501100000038","name":"NSERC","doi-asserted-by":"crossref","award":["CRDPJ532227-18"],"award-info":[{"award-number":["CRDPJ532227-18"]}],"id":[{"id":"10.13039\/501100000038","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100009192","name":"Alberta Innovates","doi-asserted-by":"publisher","award":["222300769"],"award-info":[{"award-number":["222300769"]}],"id":[{"id":"10.13039\/501100009192","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000196","name":"Canada Foundation for Innovation","doi-asserted-by":"publisher","award":["JELF grant (36605)"],"award-info":[{"award-number":["JELF grant (36605)"]}],"id":[{"id":"10.13039\/501100000196","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000038","name":"NSERC","doi-asserted-by":"crossref","award":["RTI-2021-00675"],"award-info":[{"award-number":["RTI-2021-00675"]}],"id":[{"id":"10.13039\/501100000038","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["www.ploscompbiol.org"],"crossmark-restriction":false},"short-container-title":["PLoS Comput Biol"],"abstract":"<jats:p>Machine Learning models have been frequently used in transcriptome analyses. Particularly, Representation Learning (RL), e.g., autoencoders, are effective in learning critical representations in noisy data. However, learned representations, e.g., the \u201clatent variables\u201d in an autoencoder, are difficult to interpret, not to mention prioritizing essential genes for functional follow-up. In contrast, in traditional analyses, one may identify important genes such as Differentially Expressed (DiffEx), Differentially Co-Expressed (DiffCoEx), and Hub genes. Intuitively, the complex gene-gene interactions may be beyond the capture of marginal effects (DiffEx) or correlations (DiffCoEx and Hub), indicating the need of powerful RL models. However, the lack of interpretability and individual target genes is an obstacle for RL\u2019s broad use in practice. To facilitate interpretable analysis and gene-identification using RL, we propose \u201cCritical genes\u201d, defined as genes that contribute highly to learned representations (e.g., latent variables in an autoencoder). As a proof-of-concept, supported by eXplainable Artificial Intelligence (XAI), we implemented eXplainable Autoencoder for Critical genes (XA4C) that quantifies each gene\u2019s contribution to latent variables, based on which Critical genes are prioritized. Applying XA4C to gene expression data in six cancers showed that Critical genes capture essential pathways underlying cancers. Remarkably, <jats:italic>Critical genes has little overlap with Hub or DiffEx genes<\/jats:italic>, <jats:italic>however<\/jats:italic>, <jats:italic>has a higher enrichment in a comprehensive disease gene database (DisGeNET) and a cancer-specific database (COSMIC)<\/jats:italic>, <jats:italic>evidencing its potential to disclose massive unknown biology<\/jats:italic>. As an example, we discovered five Critical genes sitting in the center of Lysine degradation (hsa00310) pathway, displaying distinct interaction patterns in tumor and normal tissues. In conclusion, XA4C facilitates explainable analysis using RL and Critical genes discovered by explainable RL empowers the study of complex interactions.<\/jats:p>","DOI":"10.1371\/journal.pcbi.1011476","type":"journal-article","created":{"date-parts":[[2023,10,2]],"date-time":"2023-10-02T17:27:51Z","timestamp":1696267671000},"page":"e1011476","update-policy":"https:\/\/doi.org\/10.1371\/journal.pcbi.corrections_policy","source":"Crossref","is-referenced-by-count":9,"title":["XA4C: eXplainable representation learning via Autoencoders revealing Critical genes"],"prefix":"10.1371","volume":"19","author":[{"given":"Qing","family":"Li","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yang","family":"Yu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Pathum","family":"Kossinna","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Theodore","family":"Lun","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Wenyuan","family":"Liao","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2701-0711","authenticated-orcid":true,"given":"Qingrun","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"340","published-online":{"date-parts":[[2023,10,2]]},"reference":[{"key":"pcbi.1011476.ref001","volume-title":"Deep learning:","author":"I Goodfellow","year":"2016"},{"issue":"3","key":"pcbi.1011476.ref002","doi-asserted-by":"crossref","first-page":"500","DOI":"10.1038\/nprot.2011.457","article-title":"Using probabilistic estimation of expression residuals (PEER) to obtain increased power and interpretability of gene expression analyses.","volume":"7","author":"O Stegle","year":"2012","journal-title":"Nat Protoc"},{"issue":"5","key":"pcbi.1011476.ref003","article-title":"MultiPLIER: A Transfer Learning Framework for Transcriptomics Reveals Systemic Features of Rare Disease","volume":"8","author":"JN Taroni","year":"2019","journal-title":"Cell Syst"},{"issue":"1","key":"pcbi.1011476.ref004","doi-asserted-by":"crossref","DOI":"10.1038\/s41467-020-14666-6","article-title":"Deriving disease modules from the compressed transcriptional space embedded in a deep autoencoder.","volume":"11","author":"SK Dwivedi","year":"2020","journal-title":"Nat Commun."},{"key":"pcbi.1011476.ref005","article-title":"Autoencoder-transformed transcriptome improves genotype-phenotype association studies.","author":"B Jiayi","year":"2023","journal-title":"bioRxiv."},{"key":"pcbi.1011476.ref006","first-page":"10","article-title":"Single-cell RNA-seq denoising using a deep count autoencoder.","author":"G Eraslan","year":"2019","journal-title":"Nat Commun."},{"key":"pcbi.1011476.ref007","first-page":"12","article-title":"Fast and precise single-cell data analysis using a hierarchical autoencoder.","author":"D Tran","year":"2021","journal-title":"Nat Commun."},{"key":"pcbi.1011476.ref008","first-page":"22","article-title":"XOmiVAE: an interpretable deep learning model for cancer classification using high-dimensional omics data","author":"E Withnell","year":"2021","journal-title":"Brief Bioinform"},{"issue":"1","key":"pcbi.1011476.ref009","article-title":"A Two-Stage Poisson Model for Testing RNA-Seq Data.","volume":"10","author":"PL Auer","year":"2011","journal-title":"Statistical Applications in Genetics and Molecular Biology."},{"issue":"4","key":"pcbi.1011476.ref010","doi-asserted-by":"crossref","first-page":"507","DOI":"10.1093\/bioinformatics\/btk005","article-title":"EDGE: extraction and analysis of differential gene expression","volume":"22","author":"JT Leek","year":"2006","journal-title":"Bioinformatics"},{"issue":"1","key":"pcbi.1011476.ref011","doi-asserted-by":"crossref","first-page":"136","DOI":"10.1093\/bioinformatics\/btp612","article-title":"DEGseq: an R package for identifying differentially expressed genes from RNA-seq data","volume":"26","author":"LK Wang","year":"2010","journal-title":"Bioinformatics"},{"key":"pcbi.1011476.ref012","doi-asserted-by":"crossref","DOI":"10.1186\/1471-2105-11-497","article-title":"DiffCoEx: a simple and sensitive method to find differentially coexpressed gene modules.","volume":"11","author":"BM Tesson","year":"2010","journal-title":"BMC Bioinformatics"},{"key":"pcbi.1011476.ref013","first-page":"9","article-title":"WGCNA: an R package for weighted correlation network analysis.","author":"P Langfelder","year":"2008","journal-title":"BMC Bioinformatics"},{"issue":"8","key":"pcbi.1011476.ref014","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pone.0135831","article-title":"CoGA: An R Package to Identify Differentially Co-Expressed Gene Sets by Analyzing the Graph Spectra.","volume":"10","author":"SD Santos","year":"2015","journal-title":"PLoS One"},{"issue":"8","key":"pcbi.1011476.ref015","doi-asserted-by":"crossref","first-page":"1121","DOI":"10.1093\/bioinformatics\/btw788","article-title":"MetaDCN: meta-analysis framework for differential co-expression network detection with an application in breast cancer","volume":"33","author":"L Zhu","year":"2017","journal-title":"Bioinformatics"},{"issue":"1","key":"pcbi.1011476.ref016","doi-asserted-by":"crossref","DOI":"10.1186\/s12859-020-03836-4","article-title":"Biological interpretation of deep neural network for phenotype prediction based on gene expression","volume":"21","author":"B Hanczar","year":"2020","journal-title":"BMC Bioinformatics"},{"key":"pcbi.1011476.ref017","doi-asserted-by":"crossref","first-page":"106619","DOI":"10.1016\/j.compbiomed.2023.106619","article-title":"Explainable artificial intelligence model for identifying COVID-19 gene biomarkers","volume":"154","author":"FH Yagin","year":"2023","journal-title":"Comput Biol Med"},{"issue":"5","key":"pcbi.1011476.ref018","doi-asserted-by":"crossref","DOI":"10.3390\/metabo13050589","article-title":"A Fecal-Microbial-Extracellular-Vesicles-Based Metabolomics Machine Learning Framework and Biomarker Discovery for Predicting Colorectal Cancer Patients.","volume":"13","author":"FH Yagin","year":"2023","journal-title":"Metabolites"},{"key":"pcbi.1011476.ref019","doi-asserted-by":"crossref","first-page":"1190429","DOI":"10.3389\/fmed.2023.1190429","article-title":"Editorial: Explainable multimodal AI in cancer patient care: how can we reduce the gap between technology and practice?","volume":"10","author":"M Rosen-Zvi","year":"2023","journal-title":"Front Med (Lausanne)"},{"issue":"37","key":"pcbi.1011476.ref020","doi-asserted-by":"crossref","DOI":"10.1126\/scirobotics.aay7120","article-title":"XAI-Explainable artificial intelligence","volume":"4","author":"D Gunning","year":"2019","journal-title":"Science Robotics"},{"key":"pcbi.1011476.ref021","first-page":"30","article-title":"A Unified Approach to Interpreting Model Predictions","author":"SM Lundberg","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"pcbi.1011476.ref022","first-page":"307","article-title":"A value for n-person games","author":"LS Shapley","year":"1953","journal-title":"Contributions to the Theory of Games II"},{"key":"pcbi.1011476.ref023","author":"S. Gillies","year":"2013","journal-title":"The Shapely user manual"},{"issue":"10","key":"pcbi.1011476.ref024","doi-asserted-by":"crossref","first-page":"1113","DOI":"10.1038\/ng.2764","article-title":"The Cancer Genome Atlas Pan-Cancer analysis project","volume":"45","author":"JN Weinstein","year":"2013","journal-title":"Nat Genet"},{"key":"pcbi.1011476.ref025","first-page":"9091","article-title":"An Introduction to Deep Learning.","author":"P Hollensen","year":"2015","journal-title":"Lect Notes Artif Int."},{"issue":"4","key":"pcbi.1011476.ref026","first-page":"1","article-title":"Xgboost: extreme gradient boosting.","volume":"1","author":"T Chen","year":"2015","journal-title":"R package version 04\u20132."},{"issue":"3","key":"pcbi.1011476.ref027","doi-asserted-by":"crossref","first-page":"263","DOI":"10.1177\/0049124105283119","article-title":"An introduction to ensemble methods for data analysis.","volume":"34","author":"RA Berk","year":"2006","journal-title":"Sociol Method Res"},{"issue":"9","key":"pcbi.1011476.ref028","doi-asserted-by":"crossref","DOI":"10.3390\/cancers10090337","article-title":"Oxidative Phosphorylation: A Target for Novel Therapeutic Strategies Against Ovarian Cancer.","volume":"10","author":"AP Nayak","year":"2018","journal-title":"Cancers (Basel)."},{"issue":"1","key":"pcbi.1011476.ref029","doi-asserted-by":"crossref","first-page":"6380","DOI":"10.1038\/s41598-020-63448-z","article-title":"High expression of oxidative phosphorylation genes predicts improved survival in squamous cell carcinomas of the head and neck and lung","volume":"10","author":"M Frederick","year":"2020","journal-title":"Sci Rep"},{"issue":"21","key":"pcbi.1011476.ref030","doi-asserted-by":"crossref","first-page":"5572","DOI":"10.1158\/0008-5472.CAN-20-3242","article-title":"Oxidative Phosphorylation Is a Metabolic Vulnerability in Chemotherapy-Resistant Triple-Negative Breast Cancer","volume":"81","author":"KW Evans","year":"2021","journal-title":"Cancer Res"},{"issue":"11","key":"pcbi.1011476.ref031","doi-asserted-by":"crossref","first-page":"2482","DOI":"10.1158\/1078-0432.CCR-17-3070","article-title":"Oxidative Phosphorylation as an Emerging Target in Cancer Therapy","volume":"24","author":"TM Ashton","year":"2018","journal-title":"Clin Cancer Res"},{"issue":"6","key":"pcbi.1011476.ref032","doi-asserted-by":"crossref","first-page":"343","DOI":"10.1002\/cbf.1149","article-title":"The role of glutathione in cancer","volume":"22","author":"GK Balendiran","year":"2004","journal-title":"Cell Biochem Funct"},{"issue":"7","key":"pcbi.1011476.ref033","doi-asserted-by":"crossref","first-page":"2291","DOI":"10.1083\/jcb.201804161","article-title":"Glutathione metabolism in cancer progression and treatment resistance","volume":"217","author":"A Bansal","year":"2018","journal-title":"J Cell Biol"},{"issue":"10","key":"pcbi.1011476.ref034","doi-asserted-by":"crossref","DOI":"10.3390\/biom10101429","article-title":"Role of Glutathione in Cancer: From Mechanisms to Therapies.","volume":"10","author":"L Kennedy","year":"2020","journal-title":"Biomolecules"},{"issue":"1","key":"pcbi.1011476.ref035","doi-asserted-by":"crossref","first-page":"262","DOI":"10.1186\/s12906-020-02981-w","article-title":"Cell metabolomics to study the function mechanism of Cyperus rotundus L. on triple-negative breast cancer cells","volume":"20","author":"S Ma","year":"2020","journal-title":"BMC Complement Med Ther"},{"issue":"12","key":"pcbi.1011476.ref036","article-title":"Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2","volume":"15","author":"MI Love","year":"2014","journal-title":"Genome Biol"},{"issue":"1\u20132","key":"pcbi.1011476.ref037","doi-asserted-by":"crossref","first-page":"119","DOI":"10.1016\/S0303-2647(98)00017-3","article-title":"Computation with the KEGG pathway database.","volume":"47","author":"H Ogata","year":"1998","journal-title":"Biosystems."},{"issue":"2","key":"pcbi.1011476.ref038","doi-asserted-by":"crossref","first-page":"318","DOI":"10.1038\/sj.bjc.6602928","article-title":"Cosmic 2005.","volume":"94","author":"S Forbes","year":"2006","journal-title":"Br J Cancer"},{"key":"pcbi.1011476.ref039","first-page":"2015","article-title":"DisGeNET: a discovery platform for the dynamical exploration of human diseases and their genes.","author":"J Pinero","year":"2015","journal-title":"Database (Oxford)."},{"key":"pcbi.1011476.ref040","article-title":"An introduction to autoencoders.","author":"U. Michelucci","year":"2022","journal-title":"arXiv."},{"key":"pcbi.1011476.ref041","first-page":"12084","article-title":"Anomaly Detection with Partitioning Overfitting Autoencoder Ensembles","author":"B Lorbeer","year":"2022","journal-title":"Proc Spie"},{"key":"pcbi.1011476.ref042","first-page":"2018","article-title":"Deep Sparse Autoencoder for Feature Extraction and Diagnosis of Locomotive Adhesion Status.","author":"CF Zhang","year":"2018","journal-title":"J Control Sci Eng."},{"issue":"5","key":"pcbi.1011476.ref043","doi-asserted-by":"crossref","first-page":"1719","DOI":"10.1007\/s13042-016-0550-y","article-title":"Research on denoising sparse autoencoder.","volume":"8","author":"LH Meng","year":"2017","journal-title":"Int J Mach Learn Cyb"},{"key":"pcbi.1011476.ref044","first-page":"785","article-title":"XGBoost: A Scalable Tree Boosting System.","author":"TQ Chen","year":"2016","journal-title":"Kdd\u201916: Proceedings of the 22nd Acm Sigkdd International Conference on Knowledge Discovery and Data Mining."},{"key":"pcbi.1011476.ref045","doi-asserted-by":"crossref","first-page":"268","DOI":"10.1007\/978-3-319-59650-1_23","article-title":"A Study on the Noise Label Influence in Boosting Algorithms: AdaBoost, GBM and XGBoost.","volume":"10334","author":"A Gomez-Rios","year":"2017","journal-title":"Hybrid Artificial Intelligent Systems, Hais 2017."},{"key":"pcbi.1011476.ref046","first-page":"2825","article-title":"Scikit-learn: Machine learning in Python.","volume":"12","author":"F Pedregosa","year":"2011","journal-title":"the Journal of machine Learning research"},{"issue":"1","key":"pcbi.1011476.ref047","first-page":"26","article-title":"Hyperparameter optimization for machine learning models based on Bayesian optimization","volume":"17","author":"J Wu","year":"2019","journal-title":"Journal of Electronic Science and Technology"},{"key":"pcbi.1011476.ref048","doi-asserted-by":"crossref","first-page":"123672","DOI":"10.1109\/ACCESS.2021.3110336","article-title":"Explainable Artificial Intelligence Based Framework for Non-Communicable Diseases Prediction.","volume":"9","author":"K Davagdorj","year":"2021","journal-title":"Ieee Access"},{"key":"pcbi.1011476.ref049","article-title":"Adam: A method for stochastic optimization.","author":"DP Kingma","year":"2014","journal-title":"arXiv"},{"key":"pcbi.1011476.ref050","first-page":"32","article-title":"PyTorch: An Imperative Style, High-Performance Deep Learning Library","author":"A Paszke","year":"2019","journal-title":"Advances in Neural Information Processing Systems"},{"key":"pcbi.1011476.ref051","volume-title":"Hands-On Gradient Boosting with XGBoost and scikit-learn: Perform accessible machine learning and extreme gradient boosting with Python","author":"C Wade","year":"2020"},{"key":"pcbi.1011476.ref052","author":"SM Lundberg","year":"2018","journal-title":"Consistent individualized feature attribution for tree ensembles"},{"issue":"4","key":"pcbi.1011476.ref053","doi-asserted-by":"crossref","first-page":"433","DOI":"10.1002\/wics.101","article-title":"Principal component analysis.","volume":"2","author":"H Abdi","year":"2010","journal-title":"Wiley interdisciplinary reviews: computational statistics."},{"issue":"18","key":"pcbi.1011476.ref054","doi-asserted-by":"crossref","first-page":"3710","DOI":"10.1093\/bioinformatics\/bth456","article-title":"GO::TermFinder\u2014open source software for accessing Gene Ontology information and finding significantly enriched Gene Ontology terms associated with a list of genes","volume":"20","author":"EI Boyle","year":"2004","journal-title":"Bioinformatics"},{"issue":"W1","key":"pcbi.1011476.ref055","doi-asserted-by":"crossref","first-page":"W199","DOI":"10.1093\/nar\/gkz401","article-title":"WebGestalt 2019: gene set analysis toolkit with revamped UIs and APIs","volume":"47","author":"YX Liao","year":"2019","journal-title":"Nucleic Acids Res"}],"updated-by":[{"DOI":"10.1371\/journal.pcbi.1011476","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2023,10,12]],"date-time":"2023-10-12T00:00:00Z","timestamp":1697068800000}}],"container-title":["PLOS Computational Biology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1011476","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,10,12]],"date-time":"2023-10-12T17:48:19Z","timestamp":1697132899000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1011476"}},"subtitle":[],"editor":[{"given":"Shihua","family":"Zhang","sequence":"first","affiliation":[],"role":[{"role":"editor","vocabulary":"crossref"}]}],"short-title":[],"issued":{"date-parts":[[2023,10,2]]},"references-count":55,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2023,10,2]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pcbi.1011476","relation":{"new_version":[{"id-type":"doi","id":"10.1371\/journal.pcbi.1011476","asserted-by":"object"}]},"ISSN":["1553-7358"],"issn-type":[{"value":"1553-7358","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,10,2]]}}}