MethylSeg v1.0.0 code
Year: 2026
Type: software
Abstract: MethylSeg is an open-source Python toolkit for identifying methylation domains in whole-genome bisulfite sequencing (WGBS) and DNA methylation microarray data. Unlike approaches that rely primarily on methylation levels at individual CpGs, MethylSeg incorporates methylation patterns across multiple genomic scales to characterize the local and regional context surrounding each CpG. The MethylSeg workflow supports data preparation, context-aware feature generation, methylation-state assignment, hidden Markov model–based genome segmentation, region cleaning, and visualization. It includes separate modeling strategies for densely sampled WGBS data and sparse, unevenly distributed microarray probes. The resulting segmentation identifies spatially coherent regions representing high-, intermediate-, low-, and PMD-associated methylation states and produces standardized BED files with accompanying region metadata. MethylSeg has been benchmarked using synthetic and experimental datasets and evaluated through comparisons with established PMD-detection methods. Its PMD calls reproduce expected biological characteristics, including H3K36me2 depletion and association with lamina-associated domains. The software enables PMD analysis in datasets for which existing approaches are limited, particularly DNA methylation microarrays and large public cohorts such as The Cancer Genome Atlas. This Zenodo record archives MethylSeg v1.0.0, the first stable release of the software and the version associated with its initial manuscript submission.
Authors: Tye, Jacob, Clement, Kendell
Institution: University of Utah
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7214035160