Expanded chromatin accessibility mapping explains genetic variation associated with complex traits in liver

Genome-wide association studies (GWAS) have identified thousands of genomic loci associated with a variety of common, complex human traits. The contribution of genetic variants to gene expression regulation has been well studied, supporting the idea that gene expression plays a causal role at some complex trait-associated loci. However, many current studies have not comprehensively investigated the impact of genetic variation on chromatin accessibility at a large scale within a single tissue. Genetic variants associated with differences in chromatin accessibility, known as chromatin accessibility quantitative trait loci (caQTLs), are major contributors to gene expression differences and GWAS signals. We assessed chromatin accessibility in 189 human liver tissue samples using ATAC-seq and identified over two million accessible chromatin regions enriched for gene regulatory characteristics. We integrated chromatin accessibility and genotype data from 175 samples and identified 14,076 caQTLs. Using publicly available blood lipid GWAS data, we found 157 loci where the colocalization of caQTLs, expression quantitative trait loci (eQTLs), and GWAS signals generated specific molecular hypotheses about causal regulatory elements, affected genes, and, in some cases, transcription factors, resolving these associations to single-nucleotide resolution. We performed a comprehensive analysis of the GWAS signals that remain without a proposed mechanism beyond liver caQTLs and eQTLs. After incorporating additional potential regulatory mechanism data, we found that approximately 26% of blood lipid GWAS signals remain without a proposed mechanism. Overall, our results demonstrate the benefits of integrating multiple datasets to improve our understanding of GWAS signals while emphasizing the need for additional experiments to fully characterize them.