Focus on single-gene effects limits discovery and interpretation of complex-trait-associated variants.

Abstract

Standard quantitative trait locus (QTL) mapping approaches consider variant effects on a single gene at a time, despite abundant evidence of allelic pleiotropy, where a single variant can affect multiple genes simultaneously. While allelic pleiotropy describes variant effects on both local and distal genes or a mixture of molecular effects on a single gene, here, we specifically investigate allelic expression "proxitropy," where a single variant influences the expression of multiple, neighboring genes. We introduce a multi-gene expression QTL (eQTL) mapping framework-cis-principal-component eQTL (cis-pc eQTL or pcQTL)-to identify variants associated with shared axes of expression variation across a cluster of neighboring genes. We perform pcQTL mapping in 13 GTEx human tissues and discover novel loci undetected by single-gene approaches. In total, we identify an average of 1,396 pcQTLs/tissue, 27% of which were not discovered by single-gene methods. These novel pcQTLs colocalized with an additional 176 genome-wide association study (GWAS) trait-associated variants and increased the number of colocalizations by 33% over single-gene QTL mapping. These findings highlight the idea that moving beyond single-gene-at-a-time approaches toward multi-gene methods can offer a more comprehensive view of gene regulation and complex-trait-associated variation.