A framework for genome-wide characterization of DNA methylation and variant effect prediction.

A harmonized compendium spanning over 90 tissues and cell types enables genome-wide discovery of context-dependent methylation classes and variant effects.

• Why it matters: Genome-wide methylation mapping across diverse cell types and tissues, and predicting how variants alter methylation in specific contexts, remains difficult.
• What they did: The framework integrates disparate methylation datasets into a unified resource, identifies methylation classes, and introduces Wreath, a deep learning sequence model for context-specific methylation impact prediction.
• The result: Using Wreath, the study characterizes how noncoding pathogenic variants affect tissue-specific methylation, supporting mechanistic hypotheses for individual regulatory variants.

AI summary of the abstract · may contain errors · not medical advice · read the paper.