PLoS Computational Biology (Jun 2020)

A framework for integrating directed and undirected annotations to build explanatory models of cis-eQTL data.

  • David Lamparter,
  • Rajat Bhatnagar,
  • Katja Hebestreit,
  • T Grant Belgard,
  • Alice Zhang,
  • Victor Hanson-Smith

DOI
https://doi.org/10.1371/journal.pcbi.1007770
Journal volume & issue
Vol. 16, no. 6
p. e1007770

Abstract

Read online

A longstanding goal of regulatory genetics is to understand how variants in genome sequences lead to changes in gene expression. Here we present a method named Bayesian Annotation Guided eQTL Analysis (BAGEA), a variational Bayes framework to model cis-eQTLs using directed and undirected genomic annotations. We used BAGEA to integrate directed genomic annotations with eQTL summary statistics from tissues of various origins. This analysis revealed epigenetic marks that are relevant for gene expression in different tissues and cell types. We estimated the predictive power of the models that were fitted based on directed genomic annotations. This analysis showed that, depending on the underlying eQTL data used, the directed genomic annotations could predict up to 1.5% of the variance observed in the expression of genes with top nominal eQTL association p-values < 10-7. For genes with estimated effect sizes in the top 25% quantile, up to 5% of the expression variance could be predicted. Based on our results, we recommend the use of BAGEA for the analysis of cis-eQTL data to reveal annotations relevant to expression biology.