{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"bioRxiv"}],"indexed":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T10:48:44Z","timestamp":1785408524558,"version":"3.56.0"},"posted":{"date-parts":[[2025,10,17]]},"group-title":"Bioinformatics","reference-count":21,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2025,10,17]],"date-time":"2025-10-17T00:00:00Z","timestamp":1760659200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.biorxiv.org\/about\/FAQ#license"}],"funder":[{"award":["2004751"],"award-info":[{"award-number":["2004751"]}],"id":[{"id":"https:\/\/ror.org\/021nxhr62","id-type":"ROR","asserted-by":"publisher"}]},{"award":["2125798"],"award-info":[{"award-number":["2125798"]}],"id":[{"id":"https:\/\/ror.org\/021nxhr62","id-type":"ROR","asserted-by":"publisher"}]},{"award":["2344169"],"award-info":[{"award-number":["2344169"]}],"id":[{"id":"https:\/\/ror.org\/021nxhr62","id-type":"ROR","asserted-by":"publisher"}]},{"award":["2319522"],"award-info":[{"award-number":["2319522"]}],"id":[{"id":"https:\/\/ror.org\/021nxhr62","id-type":"ROR","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000002","name":"National Institutes of Health","doi-asserted-by":"crossref","award":["1R01AI179686-01A1"],"award-info":[{"award-number":["1R01AI179686-01A1"]}],"id":[{"id":"10.13039\/100000002","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2025,10,17]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>Human cancer is highly heterogeneous, resulting in variable drug resistance and clinical outcomes. This complexity hinders accurate prognosis prediction and the development of targeted therapies. Molecular subtyping addresses these challenges by grouping cancers into more homogeneous subsets based on molecular characteristics, enabling subtype-specific treatment strategies. Subtyping is crucial for early diagnosis, personalized therapy, and improved survival by capturing differential therapeutic responses. Existing approaches to cancer subtyping fall into supervised and unsupervised categories. Supervised methods, often trained on The Cancer Genome Atlas (TCGA), rely on predefined subtype annotations but face limitations in generalizability and novel subtype discovery. Unsupervised methods, while capable of identifying new subtypes, may overlook widely recognized ones, hindering consistency with established classifications. Multi-omics approaches improve accuracy but are constrained by costs and data collection.<\/jats:p>\n                <jats:p>\n                  We propose CancerSubminer, a hybrid subtyping framework that integrates supervised and unsupervised learning. A subtype classifier is first trained on labeled data, after which clustering is applied to extracted features, with low-confidence samples reassigned to refine subtype boundaries. Model is retrained with the refined subtypes, and adversarial training corrects batch effects and learns domain-invariant features across labeled TCGA and unlabeled external datasets. A subsequent semi-supervised fine-tuning phase aligns subtypes between datasets and designates low-confidence samples as potential novel candidates. CancerSubminer was evaluated on five cancer types, including breast, bladder, brain, kidney, and thyroid cancers, using TCGA methylation data with annotated subtypes and unlabeled datasets from the Gene Expression Omnibus. The framework outperformed state-of-the-art subtyping models (iClusterPlus, iClusterBayes, NEMO) and clustering methods (Spectral, K-means). Kaplan\u2013Meier survival analysis demonstrated significant prognostic separation (p &lt; 0.05) for all cancers, including thyroid cancer where predefined subtypes showed no significance but CancerSubminer-derived subtypes did. These findings highlight CancerSubminer\u2019s ability to identify distinct prognostic subtypes, mitigate batch effects, and improve prognostic stratification across heterogeneous datasets. CancerSubminer is publicly available at\n                  <jats:ext-link xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" ext-link-type=\"uri\" xlink:href=\"https:\/\/github.com\/joungmin-choi\/CancerSubminer\">https:\/\/github.com\/joungmin-choi\/CancerSubminer<\/jats:ext-link>\n                  .\n                <\/jats:p>","DOI":"10.1101\/2025.10.17.682936","type":"posted-content","created":{"date-parts":[[2025,10,18]],"date-time":"2025-10-18T01:25:19Z","timestamp":1760750719000},"source":"Crossref","is-referenced-by-count":0,"title":["CancerSubminer: an integrated framework for cancer subtyping using supervised and unsupervised learning on DNA methylation profiles"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-2090-3330","authenticated-orcid":false,"given":"Joung Min","family":"Choi","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Liqing","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2025102208550694000_2025.10.17.682936v1.1","doi-asserted-by":"publisher","DOI":"10.1038\/S41576-019-0114-6"},{"key":"2025102208550694000_2025.10.17.682936v1.2","doi-asserted-by":"crossref","unstructured":"Yuxin Chen , Yuqi Wen , Chenyang Xie , Xinjian Chen , Song He , Xiaochen Bo , and Zhongnan Zhang . Mocss: Multi-omics data clustering and cancer subtyping via shared and specific representation learning. Iscience, 26(8), 2023.","DOI":"10.1016\/j.isci.2023.107378"},{"key":"2025102208550694000_2025.10.17.682936v1.3","doi-asserted-by":"crossref","first-page":"866005","DOI":"10.3389\/fgene.2022.866005","article-title":"Molecular subtyping of cancer based on distinguishing co-expression modules and machine learning","volume":"13","year":"2022","journal-title":"Frontiers in genetics"},{"key":"2025102208550694000_2025.10.17.682936v1.4","doi-asserted-by":"publisher","DOI":"10.1200\/JCO.2008.18.1370"},{"issue":"1","key":"2025102208550694000_2025.10.17.682936v1.5","first-page":"68","article-title":"Review the cancer genome atlas (tcga): an immeasurable source of knowledge","volume":"2015","year":"2015","journal-title":"Contemporary Oncology\/Wsp\u00f3\u0142czesna Onkologia"},{"key":"2025102208550694000_2025.10.17.682936v1.6","unstructured":"Qianxinz Mo and Ronglai Shen. iClusterPlus: Integrative clustering of multi-type genomic data, 2025. R package version 1.44.0."},{"key":"2025102208550694000_2025.10.17.682936v1.7","doi-asserted-by":"publisher","DOI":"10.1093\/biostatistics\/kxx017"},{"key":"2025102208550694000_2025.10.17.682936v1.8","doi-asserted-by":"publisher","DOI":"10.1093\/bioinformatics\/btz058"},{"key":"2025102208550694000_2025.10.17.682936v1.9","doi-asserted-by":"crossref","first-page":"1150","DOI":"10.3389\/fgene.2019.01150","article-title":"Dna methylation cancer biomarkers: translation to the clinic","volume":"10","year":"2019","journal-title":"Frontiers in genetics"},{"key":"2025102208550694000_2025.10.17.682936v1.10","doi-asserted-by":"crossref","first-page":"83","DOI":"10.3389\/fgene.2018.00083","article-title":"Adjusting for batch effects in dna methylation microarray data, a lesson learned","volume":"9","year":"2018","journal-title":"Frontiers in genetics"},{"key":"2025102208550694000_2025.10.17.682936v1.11","unstructured":"Yves Grandvalet and Yoshua Bengio . Semi-supervised learning by entropy minimization. Advances in neural information processing systems, 17, 2004."},{"key":"2025102208550694000_2025.10.17.682936v1.12","first-page":"896","volume-title":"In Workshop on challenges in representation learning, ICML","volume":"3","year":"2013"},{"key":"2025102208550694000_2025.10.17.682936v1.13","doi-asserted-by":"publisher","DOI":"10.1200\/JCO.2008.18.1370"},{"key":"2025102208550694000_2025.10.17.682936v1.14","doi-asserted-by":"publisher","DOI":"10.1038\/nature12965"},{"key":"2025102208550694000_2025.10.17.682936v1.15","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2015.12.028"},{"key":"2025102208550694000_2025.10.17.682936v1.16","doi-asserted-by":"publisher","DOI":"10.1016\/j.cell.2014.09.050"},{"key":"2025102208550694000_2025.10.17.682936v1.17","doi-asserted-by":"publisher","DOI":"10.1038\/nature12222"},{"key":"2025102208550694000_2025.10.17.682936v1.18","doi-asserted-by":"publisher","DOI":"10.1093\/nar\/30.1.207"},{"issue":"1","key":"2025102208550694000_2025.10.17.682936v1.19","doi-asserted-by":"crossref","first-page":"168","DOI":"10.1186\/s12859-023-05272-6","article-title":"meth-semicancer: a cancer subtype classification framework via semi-supervised learning utilizing dna methylation profiles","volume":"24","year":"2023","journal-title":"BMC bioinformatics"},{"key":"2025102208550694000_2025.10.17.682936v1.20","doi-asserted-by":"crossref","unstructured":"Joung Min Choi and Liqing Zhang . Breast cancer subtype prediction model integrating domain adaptation with semi-supervised learning on dna methylation profiles. In International Conference on Artificial Intelligence in Medicine, pages 357\u2013366. Springer, 2024.","DOI":"10.1007\/978-3-031-66538-7_35"},{"key":"2025102208550694000_2025.10.17.682936v1.21","article-title":"Umap: Uniform manifold approximation and projection for dimension reduction","year":"2018","journal-title":"arXiv preprint"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1101\/2025.10.17.682936","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T09:50:41Z","timestamp":1768470641000},"score":1,"resource":{"primary":{"URL":"http:\/\/biorxiv.org\/lookup\/doi\/10.1101\/2025.10.17.682936"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,17]]},"references-count":21,"URL":"https:\/\/doi.org\/10.1101\/2025.10.17.682936","relation":{},"subject":[],"published":{"date-parts":[[2025,10,17]]},"subtype":"preprint"}}