Predictive models of gene regulation from high-throughput epigenomics data

Area of Science:

• Epigenetics and Genomics
• Computational Biology
• Gene Regulation

Background:

• Gene expression is epigenetically regulated by multiple factors, suggesting an epigenetic code.
• High-throughput sequencing (HTS) offers tools to explore this code and model gene regulation quantitatively.
• Understanding epigenetic differences between cellular states is crucial for deciphering gene regulation.

Purpose of the Study:

• To develop a computational framework for integrating HTS epigenetic data.
• To build quantitative models relating epigenetic signals to gene expression differences.
• To investigate the nature of the epigenetic code and its associated genomic regions.

Main Methods:

• Developed a computational framework for systematic integration of HTS epigenetic data.
• Related epigenetic signals to gene expression by comparing two cellular conditions.
• Utilized ENCODE project data to build and validate predictive models.

Main Results:

• Successfully built a model predicting significant expression differences between two cell lines with high accuracy.
• Provided evidence for a degenerate epigenetic code involving multiple genic regions.
• Identified strong associations between signal changes in the 1st exon, 1st intron, and downstream of the polyadenylation site with expression regulation.
• Observed distinct epigenetic codes for intron-less and intron-containing genes.

Conclusions:

• The developed framework enables integrative analysis of epigenetic differences across cellular conditions.
• The findings support a degenerate epigenetic code, highlighting specific regulatory regions.
• The methodology is applicable to diverse biological studies, including cell differentiation and carcinogenesis.