{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,2]],"date-time":"2026-09-02T20:14:14Z","timestamp":1788380054971,"version":"build-2803163510"},"reference-count":35,"publisher":"Oxford University Press (OUP)","issue":"5","license":[{"start":{"date-parts":[[2026,9,2]],"date-time":"2026-09-02T00:00:00Z","timestamp":1788307200000},"content-version":"vor","delay-in-days":1,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"funder":[{"DOI":"10.13039\/100008247","name":"University of Otago","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100008247","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,9,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>DNA methylation alterations are early and stable hallmarks of cancer and represent promising biomarkers for non-invasive detection using circulating cell-free DNA (cfDNA). However, current computational approaches often model DNA sequence and methylation features separately and struggle to capture complex read-level methylation architecture in heterogeneous, low-signal liquid biopsy data. Here, we present DNAmBERT, a Transformer-based deep learning framework designed to jointly model DNA sequence context and read-level methylation haplotype structure from cfDNA methylation sequencing data. DNAmBERT integrates k-mer\u2013encoded DNA sequences with methylation haplotype tokens using a unified representation and masked language modelling objective, enabling context-aware learning of sequence\u2013epigenetic dependencies through self-attention. We evaluated DNAmBERT across multiple cfDNA methylation platforms (RRBS, cfRRBS, and cfMethyl-seq) and cancer types, including colorectal cancer, lung adenocarcinoma and hepatocellular carcinoma. In binary classification tasks, the model achieved high performance across platforms (AUC up to 0.99\u20131.00) and outperformed conventional machine learning and existing deep learning approaches. Aggregation of read-level predictions enabled quantitative tumour probability estimation at the sample level. Beyond binary detection, DNAmBERT supported multi-cancer and stage-aware classification, including early-stage disease, with multiclass AUC values up to 0.99. The framework further demonstrated effective cross-cancer transfer learning, maintaining robust performance under limited data availability. These results indicate that integrated sequence\u2013haplotype representation learning provides an accurate and scalable approach for cfDNA-based multi-cancer detection.<\/jats:p>","DOI":"10.1093\/bib\/bbag455","type":"journal-article","created":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T11:44:17Z","timestamp":1786535057000},"source":"Crossref","is-referenced-by-count":0,"title":["DNAmBERT: a transformer-based model for non-invasive cancer diagnosis using DNA sequence and methylation data"],"prefix":"10.1093","volume":"27","author":[{"given":"Maryam","family":"Yassi","sequence":"first","affiliation":[{"name":"Department of Mathematics and Statistics, University of Otago , Dunedin 9016 ,","place":["New Zealand"]},{"name":"Department of Pathology, Dunedin School of Medicine, University of Otago , Dunedin 9016 ,","place":["New Zealand"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mark","family":"Ezegbogu","sequence":"additional","affiliation":[{"name":"Department of Pathology, Dunedin School of Medicine, University of Otago , Dunedin 9016 ,","place":["New Zealand"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Euan J","family":"Rodger","sequence":"additional","affiliation":[{"name":"Department of Pathology, Dunedin School of Medicine, University of Otago , Dunedin 9016 ,","place":["New Zealand"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Peter","family":"Stockwell","sequence":"additional","affiliation":[{"name":"Department of Pathology, Dunedin School of Medicine, University of Otago , Dunedin 9016 ,","place":["New Zealand"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7276-2248","authenticated-orcid":false,"given":"Aniruddha","family":"Chatterjee","sequence":"additional","affiliation":[{"name":"Department of Pathology, Dunedin School of Medicine, University of Otago , Dunedin 9016 ,","place":["New Zealand"]},{"name":"Honorary Professor, UPES University , Dehradun 248007 ,","place":["India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6588-0219","authenticated-orcid":false,"given":"Matthew","family":"Parry","sequence":"additional","affiliation":[{"name":"Department of Mathematics and Statistics, University of Otago , Dunedin 9016 ,","place":["New Zealand"]},{"name":"Te P\u016bnaha Matatini Centre of Research Excellence, University of Auckland , Auckland 1010 ,","place":["New Zealand"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2026,9,2]]},"reference":[{"key":"2026090215184783400_ref1","doi-asserted-by":"publisher","first-page":"209","DOI":"10.3322\/caac.21660","article-title":"Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries","volume":"71","author":"Sung","year":"2021","journal-title":"CA Cancer J Clin"},{"key":"2026090215184783400_ref2","doi-asserted-by":"publisher","first-page":"1148","DOI":"10.1056\/NEJMra072067","article-title":"Epigenetics in cancer","volume":"358","author":"Esteller","year":"2008","journal-title":"N Engl J Med"},{"key":"2026090215184783400_ref3","doi-asserted-by":"publisher","first-page":"149","DOI":"10.1016\/j.semcancer.2017.08.004","article-title":"Epigenetic drivers of tumourigenesis and cancer metastasis","volume":"51","author":"Chatterjee","year":"2018","journal-title":"Semin Cancer Biol"},{"key":"2026090215184783400_ref4","doi-asserted-by":"publisher","first-page":"242","DOI":"10.1186\/s12935-023-03074-7","article-title":"Methylomics and cancer: the current state of methylation profiling and marker development for clinical care","volume":"23","author":"Liu","year":"2023","journal-title":"Cancer Cell Int"},{"key":"2026090215184783400_ref5","doi-asserted-by":"publisher","first-page":"1783","DOI":"10.1158\/2159-8290.CD-24-0296","article-title":"The epigenetic hallmarks of cancer","volume":"14","author":"Esteller","year":"2024","journal-title":"Cancer Discov"},{"key":"2026090215184783400_ref6","doi-asserted-by":"publisher","first-page":"253","DOI":"10.1038\/nrc1045","article-title":"The power and the promise of DNA methylation markers","volume":"3","author":"Laird","year":"2003","journal-title":"Nat Rev Cancer"},{"key":"2026090215184783400_ref7","doi-asserted-by":"publisher","first-page":"112","DOI":"10.1373\/clinchem.2014.222679","article-title":"Circulating tumor DNA as a liquid biopsy for cancer","volume":"61","author":"Heitzer","year":"2015","journal-title":"Clin Chem"},{"key":"2026090215184783400_ref8","doi-asserted-by":"publisher","first-page":"426","DOI":"10.1038\/nrc3066","article-title":"Cell-free nucleic acids as biomarkers in cancer patients","volume":"11","author":"Schwarzenbach","year":"2011","journal-title":"Nat Rev Cancer"},{"key":"2026090215184783400_ref9","doi-asserted-by":"publisher","first-page":"57","DOI":"10.1016\/j.cell.2015.11.050","article-title":"Cell-free DNA comprises an In vivo nucleosome footprint that informs its tissues-of-origin","volume":"164","author":"Snyder","year":"2016","journal-title":"Cell"},{"key":"2026090215184783400_ref10","doi-asserted-by":"publisher","first-page":"499","DOI":"10.1016\/j.molmed.2024.03.007","article-title":"Cell-free DNA methylation in the clinical management of lung cancer","volume":"30","author":"Ezegbogu","year":"2024","journal-title":"Trends Mol Med"},{"key":"2026090215184783400_ref11","doi-asserted-by":"publisher","first-page":"168","DOI":"10.1186\/s13148-023-01585-8","article-title":"Unlocking the secrets: the power of methylation-based cfDNA detection of tissue damage in organ systems","volume":"15","author":"Zhang","year":"2023","journal-title":"Clin Epigenetics"},{"key":"2026090215184783400_ref12","first-page":"150","article-title":"Clinical cancer epigenetics","volume":"17","year":"2025","journal-title":"Clin Epigenetics"},{"key":"2026090215184783400_ref13","doi-asserted-by":"publisher","first-page":"53","DOI":"10.1186\/s13059-017-1191-5","article-title":"CancerLocator: non-invasive cancer diagnosis and tissue-of-origin prediction using methylation profiles of cell-free DNA","volume":"18","author":"Kang","year":"2017","journal-title":"Genome Biol"},{"key":"2026090215184783400_ref14","doi-asserted-by":"publisher","first-page":"e89","DOI":"10.1093\/nar\/gky423","article-title":"CancerDetector: ultrasensitive and non-invasive cancer detection at the resolution of individual reads using cell-free DNA methylation sequencing data","volume":"46","author":"Li","year":"2018","journal-title":"Nucleic Acids Res"},{"key":"2026090215184783400_ref15","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbac200","article-title":"A computational framework to unify orthogonal information in DNA methylation and copy number aberrations in cell-free DNA for early cancer detection","volume":"23","author":"Wei","year":"2022","journal-title":"Brief Bioinform"},{"key":"2026090215184783400_ref16","doi-asserted-by":"publisher","first-page":"6042","DOI":"10.1038\/s41467-023-41774-w","article-title":"Multimodal analysis of cell-free DNA whole-methylome sequencing for cancer detection and localization","volume":"14","author":"Bie","year":"2023","journal-title":"Nat Commun"},{"key":"2026090215184783400_ref17","doi-asserted-by":"publisher","first-page":"9","DOI":"10.1186\/s13073-023-01280-6","article-title":"Multimodal epigenetic sequencing analysis (MESA) of cell-free DNA for non-invasive colorectal cancer detection","volume":"16","author":"Li","year":"2024","journal-title":"Genome Med"},{"key":"2026090215184783400_ref18","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbab250","article-title":"DISMIR: deep learning-based noninvasive cancer detection by integrating DNA sequence and methylation information of individual cell-free DNA reads","volume":"22","author":"Li","year":"2021","journal-title":"Brief Bioinform"},{"key":"2026090215184783400_ref19","doi-asserted-by":"publisher","first-page":"788","DOI":"10.1038\/s41467-025-55920-z","article-title":"MethylBERT enables read-level DNA methylation pattern identification and tumour deconvolution using a transformer-based model","volume":"16","author":"Jeong","year":"2025","journal-title":"Nat Commun"},{"key":"2026090215184783400_ref20","doi-asserted-by":"publisher","first-page":"2755","DOI":"10.1002\/1878-0261.13745","article-title":"Transformer-based representation learning and multiple-instance learning for cancer diagnosis exclusively from raw sequencing fragments of bisulfite-treated plasma cell-free DNA","volume":"18","author":"Liu","year":"2024","journal-title":"Mol Oncol"},{"key":"2026090215184783400_ref21","article-title":"BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding","author":"Devlin","year":"2018"},{"key":"2026090215184783400_ref22","doi-asserted-by":"publisher","first-page":"5868","DOI":"10.1093\/nar\/gki901","article-title":"Reduced representation bisulfite sequencing for comparative high-resolution DNA methylation analysis","volume":"33","author":"Meissner","year":"2005","journal-title":"Nucleic Acids Res"},{"key":"2026090215184783400_ref23","first-page":"663195","article-title":"A versatile method for circulating cell-free DNA methylome profiling by reduced representation bisulfite sequencing","author":"De Koker","year":"2019","journal-title":"bioRxiv"},{"key":"2026090215184783400_ref24","doi-asserted-by":"publisher","first-page":"5566","DOI":"10.1038\/s41467-022-32995-6","article-title":"Cost-effective methylome sequencing of cell-free DNA for accurately detecting and locating cancer","volume":"13","author":"Stackpole","year":"2022","journal-title":"Nat Commun"},{"key":"2026090215184783400_ref25","doi-asserted-by":"publisher","first-page":"1338","DOI":"10.12688\/f1000research.15931.1","article-title":"FastQ screen: a tool for multi-genome mapping and quality control","volume":"7","author":"Wingett","year":"2018","journal-title":"F1000Res"},{"key":"2026090215184783400_ref26","author":"Krueger"},{"key":"2026090215184783400_ref27","doi-asserted-by":"publisher","first-page":"2078","DOI":"10.1093\/bioinformatics\/btp352","article-title":"The sequence alignment\/map format and SAMtools","volume":"25","author":"Li","year":"2009","journal-title":"Bioinformatics"},{"key":"2026090215184783400_ref28","doi-asserted-by":"publisher","volume":"49","journal-title":"Nat Genet","DOI":"10.1038\/ng.3805"},{"key":"2026090215184783400_ref29","doi-asserted-by":"publisher","first-page":"179","DOI":"10.1111\/j.1469-1809.1936.tb02137.x","article-title":"The use of multiple measurements in taxonomic problems","volume":"7","author":"Fisher","year":"1936","journal-title":"Ann Eugenics"},{"key":"2026090215184783400_ref30","article-title":"Computational and theoretical analysis of null space and orthogonal linear discriminant analysis","volume":"7","author":"Ye","year":"2006","journal-title":"J Mach Learn Res"},{"key":"2026090215184783400_ref31","doi-asserted-by":"publisher","first-page":"066138","DOI":"10.1103\/PhysRevE.69.066138","article-title":"Estimating mutual information","volume":"69","author":"Kraskov","year":"2004","journal-title":"Physical Review E\u2014Statistical, Nonlinear, and Soft Matter Physics"},{"key":"2026090215184783400_ref32","doi-asserted-by":"publisher","first-page":"366","DOI":"10.1016\/j.ygeno.2017.12.006","article-title":"DMRFusion: a differentially methylated region detection tool based on the ranked fusion method","volume":"110","author":"Yassi","year":"2018","journal-title":"Genomics"},{"key":"2026090215184783400_ref33","doi-asserted-by":"publisher","first-page":"4251","DOI":"10.1093\/bioinformatics\/btab413","article-title":"cfDNApipe: a comprehensive quality control and analysis pipeline for cell-free DNA high-throughput sequencing data","volume":"37","author":"Zhang","year":"2021","journal-title":"Bioinformatics"},{"key":"2026090215184783400_ref34","doi-asserted-by":"publisher","first-page":"giae102","DOI":"10.1093\/gigascience\/giae102","article-title":"cfDNA UniFlow: a unified preprocessing pipeline for cell-free DNA data from liquid biopsies","volume":"13","author":"R\u00f6ner","year":"2024","journal-title":"GigaScience"},{"key":"2026090215184783400_ref35","doi-asserted-by":"publisher","volume":"26","journal-title":"iScience","DOI":"10.1016\/j.isci.2023.106986"}],"container-title":["Briefings in Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bib\/article-pdf\/27\/5\/bbag455\/70925523\/bbag455.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bib\/article-pdf\/27\/5\/bbag455\/70925523\/bbag455.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,2]],"date-time":"2026-09-02T19:18:54Z","timestamp":1788376734000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bib\/article\/doi\/10.1093\/bib\/bbag455\/8780345"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,1]]},"references-count":35,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2026,9,1]]}},"URL":"https:\/\/doi.org\/10.1093\/bib\/bbag455","relation":{},"ISSN":["1467-5463","1477-4054"],"issn-type":[{"value":"1467-5463","type":"print"},{"value":"1477-4054","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2026,9]]},"published":{"date-parts":[[2026,9,1]]},"article-number":"bbag455"}}