Discovering high-resolution patterns of differential DNA methylation that correlate with gene expression changes

Area of Science:

• Epigenetics and Genomics
• Molecular Biology
• Bioinformatics

Background:

• DNA methylation at CpG islands typically silences gene expression.
• Genome-wide studies show modest correlations between promoter methylation and gene expression.
• Existing methods may oversimplify methylation patterns, missing complex associations.

Purpose of the Study:

• To develop a novel approach for identifying diverse differential DNA methylation patterns associated with gene expression changes.
• To investigate methylation patterns beyond traditional promoter CpG islands, including shores and hypomethylated domains.
• To improve the accuracy and comprehensiveness of methylation-expression association analyses.

Main Methods:

• Representing differential methylation as interpolated curves (signatures) from high-resolution genome-wide data.
• Clustering genes with similar methylation signature shapes and corresponding expression changes.
• Applying the method to embryonic stem cell and cancer datasets.

Main Results:

• Uncovered a diverse range of methylation patterns associated with gene expression changes.
• Identified strong associations between these novel methylation signatures and gene expression.
• Demonstrated conservation of these patterns across different cell types and disease contexts.
• An extended method outperformed existing approaches in identifying methylation-expression associations.

Conclusions:

• The developed signature-based approach captures complex DNA methylation patterns linked to gene expression.
• This method provides a more nuanced understanding of epigenetic regulation compared to traditional analyses.
• The findings have implications for understanding gene silencing in development and diseases like cancer.