Inferring causal relationships among different histone modifications and gene expression

Area of Science:

• Epigenetics and Genomics
• Computational Biology
• Molecular Biology

Background:

• Histone modifications are key epigenetic regulators of gene expression.
• These modifications are crucial for stem cell pluripotency and cancer development.
• Complex interactions, termed "histone codes," govern gene regulation.

Purpose of the Study:

• To build a Bayesian network for inferring relationships among histone modifications and gene expression.
• To analyze high-resolution genome maps from ChIP-chip and ChIP-seq data.
• To decipher complex "histone codes" and identify novel epigenetic interactions.

Main Methods:

• Utilized high-throughput ChIP-chip and ChIP-seq data for human genome-wide histone modification mapping.
• Constructed a Bayesian network to model combinatorial and causal relationships.
• Validated the network using pilot studies on Polycomb group genes and H3K27 trimethylation, and through cross-validation.

Main Results:

• A pilot network for Polycomb group genes and H3K27 trimethylation was accurately supported by existing literature.
• The developed network confirmed known relationships (e.g., H3K27me3 to silencing, H3K4me3 to activation, bivalent modifications).
• Identified novel relationships, predicting new epigenetic interactions crucial for gene regulation.

Conclusions:

• The Bayesian network approach effectively infers complex histone codes and epigenetic interactions.
• This method provides a valuable tool for analyzing ChIP-chip and ChIP-seq data.
• The findings advance our understanding of epigenetic gene regulation in stem cell pluripotency and cancer.