{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T06:03:45Z","timestamp":1784527425982,"version":"3.55.0"},"reference-count":34,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T00:00:00Z","timestamp":1784505600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>Patient subtyping is significantly challenged by intra-sample heterogeneity, which limits the effectiveness of traditional multi-class classification approaches enforcing mutually exclusive labels. Despite recent promising results in the transition from multi-class to multi-label classification, this process is not straightforward and proves hard to systemize, especially when working with datasets of small dimensions. Here, we design a novel approach, implemented in a computational framework, leveraging machine learning models and an original nested conformal prediction strategy for small datasets to enable a rigorous transition to multi-label classification. Conformal prediction can indeed offer a statistically sound approach for robust subtype predictions; yet, so far, it has been limited to be exploited only in large datasets, not often available in real biomedical scenarios. We evaluate our approach using breast cancer patient RNA-seq data from The Cancer Genome Atlas, with subtypes originally defined via the PAM50 classifier. From a methodological standpoint, our innovative nested strategy enables the successful application of conformal prediction, which commonly requires large datasets, to a relatively small sample size. Specifically, conformal prediction is adapted and applied both to the reference PAM50 classification and to multiple machine learning models to test coherence and compare the obtained multi-label predictions. Our results reveal subtype- and sample-specific complexity and highlight consistent patterns of ambiguity across models. Importantly, we distinguish between robust, inherently heterogeneous, and weak assignments, providing deeper insight into clear subtype confirmations and complexity, but also critical incoherent predictions. Thus, overall, this work demonstrates the utility of our framework for uncertainty-aware sample subtyping, particularly in small and highly unbalanced sample settings.<\/jats:p>","DOI":"10.3389\/frai.2026.1844254","type":"journal-article","created":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T05:35:43Z","timestamp":1784525743000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Novel nested conformal prediction analysis to unravel complexity in patient subtyping"],"prefix":"10.3389","volume":"9","author":[{"given":"Linda","family":"Maldera","sequence":"first","affiliation":[{"name":"Politecnico di Milano, Dipartimento di Elettronica, Informazione e Bioingengeria","place":["Milan, Italy"]},{"name":"Computational Multi-Omics of Neurological Disorders (MIND) Lab Joint Research Platform, Fondazione IRCCS Istituto Neurologico Carlo Besta","place":["Milan, Italy"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marco","family":"Masseroli","sequence":"additional","affiliation":[{"name":"Politecnico di Milano, Dipartimento di Elettronica, Informazione e Bioingengeria","place":["Milan, Italy"]},{"name":"Computational Multi-Omics of Neurological Disorders (MIND) Lab Joint Research Platform, Fondazione IRCCS Istituto Neurologico Carlo Besta","place":["Milan, Italy"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Silvia","family":"Cascianelli","sequence":"additional","affiliation":[{"name":"Politecnico di Milano, Dipartimento di Elettronica, Informazione e Bioingengeria","place":["Milan, Italy"]},{"name":"Computational Multi-Omics of Neurological Disorders (MIND) Lab Joint Research Platform, Fondazione IRCCS Istituto Neurologico Carlo Besta","place":["Milan, Italy"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,7,20]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"2794","DOI":"10.1200\/JCO.2013.54.1870","article-title":"Luminal B breast cancer: molecular characterization, clinical management, and future perspectives","volume":"32","author":"Ades","year":"2014","journal-title":"J. Clin. Oncol."},{"key":"B2","doi-asserted-by":"publisher","first-page":"106743","DOI":"10.1016\/j.bspc.2024.10674","article-title":"A reliable breast cancer diagnosis approach using an optimized deep learning and conformal prediction","volume":"98","author":"Almaslukh","year":"2024","journal-title":"Biomed. Signal Process. Control"},{"key":"B3","doi-asserted-by":"publisher","first-page":"494","DOI":"10.1561\/2200000101","article-title":"A gentle introduction to conformal prediction and distribution-free uncertainty quantification","volume":"16","author":"Angelopoulos","year":"2022","journal-title":"Found. Trends Mach. Learn"},{"key":"B4","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s13073-022-01105-y","article-title":"A novel molecular signature identifies mixed subtypes in renal cell carcinoma with poor prognosis and independent response to immunotherapy","volume":"14","author":"B\u00fcttner","year":"2022","journal-title":"Genome Med"},{"key":"B5","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s13073-023-01176-5","article-title":"Multi-label transcriptional classification of colorectal cancer reflects tumor cell population heterogeneity","volume":"15","author":"Cascianelli","year":"2023","journal-title":"Genome Med"},{"key":"B6","doi-asserted-by":"publisher","first-page":"104817","DOI":"10.1016\/j.jbi.2025.104817","article-title":"A novel machine learning-based workflow to capture intra-patient heterogeneity through transcriptional multi-label characterization and clinically relevant classification","volume":"166","author":"Cascianelli","year":"2025","journal-title":"J. Biomed. Inform"},{"key":"B7","doi-asserted-by":"publisher","first-page":"14071","DOI":"10.1038\/s41598-020-70832-2","article-title":"Machine learning for rna sequencing-based intrinsic subtyping of breast cancer","volume":"10","author":"Cascianelli","year":"2020","journal-title":"Sci. Rep"},{"key":"B8","doi-asserted-by":"publisher","first-page":"506","DOI":"10.1016\/j.cell.2015.09.033","article-title":"Comprehensive molecular portraits of invasive lobular breast cancer","volume":"163","author":"Ciriello","year":"2015","journal-title":"Cell"},{"key":"B9","first-page":"2929","article-title":"Breast cancer intrinsic subtype classification, clinical use and future trends","volume":"5","author":"Dai","year":"2015","journal-title":"Am. J. Cancer Res"},{"key":"B10","doi-asserted-by":"publisher","first-page":"245","DOI":"10.1007\/s13748-012-0021-y","article-title":"Conformal predictors in early diagnostics of ovarian and breast cancers","volume":"1","author":"Devetyarov","year":"2012","journal-title":"Progress Artif. Intell"},{"key":"B11","doi-asserted-by":"publisher","first-page":"130","DOI":"10.5858\/134.1.130","article-title":"Basal-like breast carcinoma: a phenotypically distinct entity","volume":"134","author":"Haupt","year":"2010","journal-title":"Arch. Pathol. Lab. Med"},{"key":"B12","doi-asserted-by":"publisher","first-page":"3708","DOI":"10.1158\/0008-5472.CAN-16-2574","article-title":"Assessment of breast cancer risk factors reveals subtype heterogeneity","volume":"77","author":"Holm","year":"2017","journal-title":"Cancer Res"},{"key":"B13","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s13059-018-1511-4","article-title":"Continuity of transcriptomes among colorectal cancer subtypes based on meta-analysis","volume":"19","author":"Ma","year":"2018","journal-title":"Genome Biol"},{"key":"B14","doi-asserted-by":"publisher","first-page":"4768","DOI":"10.1158\/1078-0432.CCR-21-0529","article-title":"Intratumor CMS heterogeneity impacts patient prognosis in localized colon cancer","volume":"27","author":"Marisa","year":"2021","journal-title":"Clin. Cancer Res"},{"key":"B15","doi-asserted-by":"publisher","first-page":"682","DOI":"10.1016\/B978-0-323-95502-7.00088-9","article-title":"\u201cPerformance measures for multi-class classification,\u201d","author":"Mongardi","year":"2025","journal-title":"Encyclopedia of Bioinformatics and Computational Biology"},{"key":"B16","doi-asserted-by":"publisher","first-page":"621","DOI":"10.1586\/erm.12.46","article-title":"How many molecular subtypes? Implications of the unique tumor principle in personalized medicine","volume":"12","author":"Ogino","year":"2012","journal-title":"Expert Rev. Mol. Diagn"},{"key":"B17","doi-asserted-by":"publisher","first-page":"7761","DOI":"10.1038\/s41467-022-34945-8","article-title":"Estimating diagnostic uncertainty in artificial intelligence assisted pathology using conformal prediction","volume":"13","author":"Olsson","year":"2022","journal-title":"Nat. Commun"},{"key":"B18","doi-asserted-by":"publisher","first-page":"1160","DOI":"10.1200\/JCO.2008.18.1370","article-title":"Supervised risk predictor of breast cancer based on intrinsic subtypes","volume":"27","author":"Parker","year":"2009","journal-title":"J. Clin. Oncol"},{"key":"B19","doi-asserted-by":"publisher","first-page":"1396","DOI":"10.1126\/science.1254257","article-title":"Single-cell RNA-seq highlights intratumoral heterogeneity in primary glioblastoma","volume":"344","author":"Patel","year":"2014","journal-title":"Science"},{"key":"B20","doi-asserted-by":"publisher","first-page":"7956","DOI":"10.1038\/s41598-019-44339-4","article-title":"Pca-pam50 improves consistency between breast cancer intrinsic and clinical subtyping reclassifying a subset of luminal a tumors as luminal b","volume":"9","author":"Perez","year":"2019","journal-title":"BMC Cancer"},{"key":"B21","doi-asserted-by":"publisher","first-page":"747","DOI":"10.1038\/35021093","article-title":"Molecular portraits of human breast tumours","volume":"406","author":"Perou","year":"2000","journal-title":"Nature"},{"key":"B22","doi-asserted-by":"publisher","first-page":"51645","DOI":"10.1109\/ACCESS.2025.3552934","article-title":"Conformal prediction for uncertainty quantification and reliable her2 status classification in breast cancer ihc images","volume":"13","author":"Pintawong","year":"2025","journal-title":"IEEE Access"},{"key":"B23","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1016\/j.molonc.2010.11.003","article-title":"Deconstructing the molecular portraits of breast cancer","volume":"5","author":"Prat","year":"2011","journal-title":"Mol. Oncol"},{"key":"B24","doi-asserted-by":"publisher","first-page":"218","DOI":"10.1158\/1078-0432.CCR-15-0529","article-title":"Prognostic and chemotherapy predictive value of gene-expression phenotypes in primary lung adenocarcinoma","volume":"22","author":"Ringn\u00e9r","year":"2016","journal-title":"Clin. Cancer Res"},{"key":"B25","doi-asserted-by":"publisher","first-page":"194","DOI":"10.1007\/s12672-025-04359-1","article-title":"Graph neural network-based multi-label classification of breast cancer using multiparametric mri and clinical features","volume":"17","author":"Salehzehi","year":"2025","journal-title":"Discover Oncol"},{"key":"B26","doi-asserted-by":"publisher","first-page":"66","DOI":"10.1186\/1755-8794-5-66","article-title":"Subtypes of primary colorectal tumors correlate with response to targeted treatment in colorectal cell lines","volume":"5","author":"Schlicker","year":"2012","journal-title":"BMC Med. Genomics"},{"key":"B27","doi-asserted-by":"publisher","first-page":"10869","DOI":"10.1073\/pnas.191367098","article-title":"Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications","volume":"98","author":"S\u00f8rlie","year":"2001","journal-title":"Proc. Natl. Acad. Sci. U. S. A"},{"key":"B28","doi-asserted-by":"crossref","first-page":"259","DOI":"10.1016\/0169-7439(89)80095-4","article-title":"Analysis of variance (anova)","volume":"6","author":"St","year":"1989","journal-title":"Chemometr. Intell. Lab. Syst"},{"key":"B29","doi-asserted-by":"publisher","first-page":"1","DOI":"10.4018\/jdwm.2007070101","article-title":"Multi-label classification: an overview","volume":"3","author":"Tsoumakas","year":"2009","journal-title":"Int. J. Data Warehous. Min"},{"key":"B30","doi-asserted-by":"publisher","first-page":"241","DOI":"10.1007\/s41666-021-00113-8","article-title":"Conformal prediction in clinical medical sciences","volume":"6","author":"Vazquez","year":"2022","journal-title":"J. Healthc. Inform. Res"},{"key":"B31","doi-asserted-by":"publisher","first-page":"54","DOI":"10.1186\/s12920-015-0129-6","article-title":"Development and verification of the pam50-based prosigna breast cancer gene signature assay","volume":"8","author":"Wallden","year":"2015","journal-title":"BMC Med. Genomics"},{"key":"B32","doi-asserted-by":"publisher","first-page":"42","DOI":"10.1016\/j.ccell.2017.06.003","article-title":"Tumor evolution of glioma-intrinsic gene expression subtypes associates with immunological changes in the microenvironment","volume":"32","author":"Wang","year":"2017","journal-title":"Cancer Cell"},{"key":"B33","doi-asserted-by":"publisher","first-page":"1113","DOI":"10.1038\/ng.2764","article-title":"The Cancer Genome Atlas pan-cancer analysis project","volume":"45","author":"Weinstein","year":"2013","journal-title":"Nat. Genet"},{"key":"B34","doi-asserted-by":"publisher","first-page":"btaf444","DOI":"10.1093\/bioinformatics\/btaf444","article-title":"Hallmarkgraph: a cancer hallmark informed graph neural network for classifying hierarchical tumor subtypes","volume":"41","author":"Zhang","year":"2025","journal-title":"Bioinformatics"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1844254\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T05:35:44Z","timestamp":1784525744000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1844254\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,20]]},"references-count":34,"alternative-id":["10.3389\/frai.2026.1844254"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1844254","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,20]]},"article-number":"1844254"}}