Multiple network algorithm for epigenetic modules via the integration of genome-wide DNA methylation and gene...

Area of Science:

• Epigenetics
• Cancer Genomics
• Bioinformatics

Background:

• Increasing DNA methylation and gene expression data enable extensive investigation of cancer's epigenetic mechanisms.
• Current methods integrate methylation and expression data by assuming anti-correlation, which is often unknown and difficult to determine.

Purpose of the Study:

• To develop a novel framework for analyzing epigenetic modules by simultaneously integrating DNA methylation and gene expression data.
• To address the challenge of unknown correlations between methylation and expression in network construction.

Main Methods:

• Introduction of the Epigenetic Module based on Differential Networks (EMDN) algorithm, a multiple network framework.
• Simultaneous analysis of DNA methylation and gene expression data without pre-specifying their correlation.
• Application of the EMDN algorithm to The Cancer Genome Atlas (TCGA) breast cancer data.

Main Results:

• The EMDN algorithm accurately identifies both positively and negatively correlated epigenetic modules.
• EMDN-identified modules show significantly higher enrichment in known pathways compared to other algorithms.
• These modules serve as effective biomarkers for predicting breast cancer subtypes using methylation profiles.
• Positively and negatively correlated modules are equally important for cancer subtype classification.
• Epigenetic modules derived from EMDN can estimate patient survival time, crucial for cancer therapy.

Conclusions:

• The EMDN algorithm offers an effective approach for the integrative analysis of DNA methylation and gene expression data.
• The developed model and algorithm provide a valuable tool for cancer research and biomarker discovery.
• The EMDN algorithm is available as an open-source R-package for broader accessibility.