Toward breaking the histone code: bayesian graphical models for histone modifications.

Area of Science:

• Epigenetics and Genomics
• Computational Biology
• Molecular Biology

Background:

• Histones are core components of nucleosomes, organizing DNA.
• Histone modifications (HMs) are post-translational changes on histone tails.
• Combinatorial patterns of HMs, or histone codes, regulate biological processes.

Purpose of the Study:

• To decode the complex biological network of histone modifications within specific genomic regions.
• To demonstrate how histone modification networks vary across different regulatory genomic regions.
• To establish a link between histone network attributes and genomic functions.

Main Methods:

• Development of a Bayesian graphical model for probabilistic inference of HM dependence patterns.
• Utilizing posterior inference to compute probabilities of distinct HM dependence patterns via graphs.
• Application of differential network analysis to compare HM networks across different genomic locations.

Main Results:

• The Bayesian model successfully inferred histone modification networks from ChIP-Seq data.
• Differential network analyses provided insights into the coregulation of HMs in various genomic regions.
• The study confirmed existing findings and generated new hypotheses regarding histone codes.

Conclusions:

• Bayesian graphical models offer a powerful framework for inferring histone networks.
• Borrowing strength across conditions enhances the power to detect differences in histone networks.
• This approach links histone modification patterns to genomic functions.