ipDMR

ipDMR is an R-based method that identifies differentially methylated regions (DMRs) by using auto-correlated P-values from individual CpG sites to summarize adjacent CpG signals and delineate DMR boundaries for epigenome-wide association studies on array and bisulfite sequencing data.

Key Features:

• Auto-correlated P-values: Uses auto-correlation among P-values from individual CpG sites to enhance detection of regional methylation signals.
• Peak summarization and extension: Summarizes P-values of adjacent CpGs to identify significant association peaks and extends these peaks to define DMR start and end points.
• Input compatibility: Accepts BED-format files and P-values derived from epigenome-wide association studies on array and bisulfite sequencing data.
• Performance: Simulations on real data report slightly higher true positive rates and significantly lower false discovery rates compared with existing methods.

Scientific Applications:

• Epigenome-wide association studies (EWAS): Identifies genomic regions with differential methylation from EWAS P-values for association with phenotypes or exposures.
• Gene regulation analyses: Provides precise DMR boundaries to support studies of methylation-mediated gene regulation.
• Disease mechanism research (including cancer): Detects aberrant methylation regions relevant to disease mechanisms and biomarker discovery in contexts such as cancer.

Methodology:

Computes auto-correlated P-values from individual CpG sites, summarizes P-values across adjacent CpGs to detect peaks, and extends identified peaks to delineate DMR start and end coordinates; accepts BED-format input.