methyl-eQTL

methyl-eQTL identifies gene-level relationships between DNA methylation and gene expression by mapping methylation-expression quantitative trait loci (methyl-eQTLs) to characterize epigenetic regulation in cancer.

Key Features:

• Sequential Penalized Regression Approach: Employs a sequential penalized regression method to select a sparse set of CpG loci that explain gene expression and to produce weights indicating the direction and strength of each association.
• Gene-Level Analysis: Focuses on individual CpG sites, including intergenic and distal CpGs, rather than relying on aggregate gene-level methylation summaries.
• Tissue-Specific Insights: Identifies genes strongly regulated by methylation in a tissue-specific manner, with examples including colorectal, breast, and pancreatic cancers.
• Model Validation: Models have been built and validated using data from TCGA (The Cancer Genome Atlas) and MD Anderson colorectal cohorts, demonstrating improved explanation of expression variability relative to existing methods.
• Comprehensive Resource: Produces gene-level methylation summaries that are maximally correlated with gene expression for use in integrative genomics analyses.

Scientific Applications:

• Cancer Epigenetics: Mapping methylation-driven regulation of gene expression to study tumorigenesis and prioritize potential therapeutic targets across cancer types.
• Tissue-Specific Regulatory Discovery: Identifying tissue-dependent epigenetic regulatory mechanisms by generating tissue-specific methylation–expression summaries.

Methodology:

The approach applies sequential penalized regression to identify CpG loci associated with gene expression (methyl-eQTLs) and to estimate weights that quantify the direction and strength of each CpG's effect on expression.