Reverse regression increases power for detecting trans-eQTLs

Abstract

Trans-acting expression quantitative trait loci (trans-eQTLs) are genetic variants affecting the expression of distant genes. They account for ≥70% expression heritability and could therefore facilitate uncovering mechansisms underlying the origination of complex diseases. However, unlike cis-eQTLs, identifying trans-eQTLs is challenging because of small effect sizes, tissue-specificity, and the severe multiple-testing burden. Trans-eQTLs affect multiple target genes, but aggregating evidence over individual SNP-gene associations is hampered by strong gene expression correlations resulting in correlated p-values. Our method Tejaas predicts trans-eQTLs by performing L2-regularized ‘reverse’ multiple regression of each SNP on all genes, aggregating evidence from many small trans-effects while being unaffected by the strong expression correlations. Combined with a novel non-linear, unsupervised k-nearest-neighbor method to remove confounders, Tejaas predicted 18851 unique trans-eQTLs across 49 tissues from GTEx. They are enriched in open chromatin, enhancers and other regulatory regions. Many overlap with disease-associated SNPs, pointing to tissue-specific transcriptional regulation mechanisms. Tejaas is available under GPL at https://github.com/soedinglab/tejaas.