The MethFinder web server provides an intuitive interface for exploring and predicting cancer-related DNA methylation events through three integrated modules:

Module 1 – Cancer-context methylation prediction
Predicts whether a given CpG site can undergo methylation in the cancer context based solely on its local DNA sequence. This module evaluates methylation potential and serves as the foundation for subsequent driver methylation analysis.

Module 2 – Cancer-type-specific driver methylation prediction
For methylated CpG sites, this module estimates whether they function as driver methylation events in specific cancer types, enabling the identification of driver CpG sites at single-base resolution.

Full MethFinder – Integrated driver methylation analysis
Combines both modules into a unified workflow for end-to-end prediction of driver methylations directly from DNA sequence. Users can submit DNA sequences or genomic coordinates to obtain comprehensive predictions, including site-level driver scores, associated genes, and mechanistic classification.

Currently, MethFinder supports predictions across 16 cancer types. Detailed usage instructions and supported cancer types can be found in the Readme file.

Source code: https://github.com/yuyuan12138/MethFinder
Example files: MethFinder_example.zip