Bayesian integration of genetics and epigenetics detects causal regulatory SNPs underlying expression variability

Area of Science:

• Genomics
• Systems Biology
• Bioinformatics

Background:

• Standard expression quantitative trait loci (eQTL) analysis identifies associations between genetic polymorphisms and gene expression but does not establish causality.
• Understanding the causal genetic variants that regulate gene expression is crucial for deciphering complex biological processes and disease mechanisms.

Purpose of the Study:

• To introduce eQTeL, a novel coupled Bayesian regression approach that integrates epigenetic data to identify causal regulatory single-nucleotide polymorphisms (SNPs) influencing gene expression.
• To evaluate the performance of eQTeL in explaining gene expression variance and predicting gene expression compared to existing methods.

Main Methods:

• Developed eQTeL, a coupled Bayesian regression model leveraging epigenetic data to estimate regulatory and gene interaction potential.
• Applied eQTeL to human heart data and realistic simulated datasets to assess its ability to identify causal regulatory SNPs and explain expression variance.
• Utilized functional genomics data, including allele-specific protein binding and histone modifications, to validate the biological relevance of detected SNPs.

Main Results:

• eQTeL explained a significantly greater proportion of gene expression variance and demonstrated higher prediction accuracy on human heart data compared to other methods.
• eQTeL accurately identified causal regulatory SNPs, including those with small effect sizes, in simulated data.
• SNPs identified by eQTeL were enriched for allele-specific protein binding and histone modifications, suggesting disruption of core cardiac transcription factor binding and proximity to target genes.

Conclusions:

• eQTeL effectively identifies causal regulatory SNPs by integrating epigenetic information, offering a more powerful approach than standard eQTL analysis.
• The identified SNPs capture a substantial portion of the genetic determinants of gene expression, with an estimated 58% being putatively causal.
• eQTeL provides a robust framework for dissecting the genetic architecture of gene expression and its regulatory mechanisms.