MethylBERT: A Transformer-based model for read-level DNA methylation pattern identification and tumour deconvolution

Abstract

DNA methylation (DNAm) is a key epigenetic mark that shows profound alterations in cancer. Although read-level methylomes enable more in-depth DNAm analysis due to the broad coverage and preservation of rare cell-type signals, the majority of published DNAm analysis methods have targeted array-based data such as EPIC/450K array. Here, we propose MethylBERT, a novel Transformer-based read-level methylation pattern classification model. MethylBERT identifies tumour-derived sequence reads based on their methylation patterns and genomic sequence using a Bidirectional Encoder Representations from Transformers (BERT) model. Based on the calculated classification probability, the method estimates tumour cell fractions within bulk samples and provides an assessment of the model precision. In our evaluation, MethylBERT outperforms existing deconvolution methods and demonstrates high accuracy regardless of methylation pattern complexity, read length and read coverage. Moreover, we show its potential for accurate non-invasive early cancer diagnostics by applying MethylBERT to liquid biopsy samples collected from cancer patients. MethylBERT represents a significant advancement in read-level methylome analysis. It will increase the accuracy of tumour deconvolution and enhance circulating tumour DNA studies.