Comprehensive Multiple eQTL Detection and Its Application to GWAS Interpretation

Abstract

Expression QTL (eQTL) detection has emerged as an important tool for unraveling the relationship between genetic risk factors and disease or clinical phenotypes. Most studies are predicated on the assumption that only a single causal variant explains the association signal in each interval. This greatly simplifies the statistical modeling, but is liable to biases in scenarios where multiple local causal-variants are responsible. Here, our primary goal was to address the prevalence of secondary cis-eQTL signals regulating peripheral blood gene expression locally, utilizing two large human cohort studies, each >2500 samples with accompanying whole genome genotypes. The CAGE (Consortium for the Architecture of Gene Expression) dataset is a compendium of Illumina microarray studies, and the Framingham Heart Study is a two-generation Affymetrix dataset. We also describe Bayesian colocalization analysis of the extent of sharing of cis-eQTL detected in both studies as well as with the BIOS RNAseq dataset. Stepwise conditional modeling demonstrates that multiple eQTL signals are present for ∼40% of over 3500 eGenes in both microarray datasets, and that the number of loci with additional signals reduces by approximately two-thirds with each conditioning step. Although <20% of the peak signals across platforms fine map to the same credible interval, the colocalization analysis finds that as many as 50-60% of the primary eQTL are actually shared. Subsequently, colocalization of eQTL signals with GWAS hits detected 1349 genes whose expression in peripheral blood is associated with 591 human phenotype traits or diseases, including enrichment for genes with regulatory functions. At least 10%, and possibly as many as 40%, of eQTL-trait colocalized signals are due to nonprimary cis-eQTL peaks, but just one-quarter of these colocalization signals replicated across the gene expression datasets. Our results are provided as a web-based resource for visualization of multi-site regulation of gene expression and its association with human complex traits and disease states.

Keywords: PolyQTL; colocalization; conditional association; fine mapping; gene regulation; linkage disequilibrium.

Figure 1

Discovery of eQTL in CAGE and FHS. (A) Counts of independent cis-eQTL detected in the CAGE and FHS cohorts. Red (CAGE) and blue (FHS) bars indicate the number of primary, secondary, tertiary, etc., eQTL detected conditional on the prior peaks at P < 10⁻⁵ in each sequential step. (B) Directional consistency of primary eQTL effects is indicated by plotting the magnitude and sign of eQTL effects (βs) in the validation study (y-axis) against discovery dataset (x-axis) for CAGE (left) or FHS (right).

Figure 2

Example of shared cis-eQTL signals, at the ORMDL3 locus in CAGE (A) and FHS (B). In both studies, two independent cis-eQTL were detected. rs12936231 and rs8067378 are the respective peak eSNPs for a credible interval of ∼50 SNPs, and are in complete LD (r² = 1), whereas the peak conditional secondary association is at rs17608925 in both studies.

Figure 3

Example of rank-changed cis-eQTL signals, at the JAZF1 locus in CAGE (A) and FHS (B). For CAGE, rs2158799 and rs498475 are the peak eSNPs in two independent credible intervals, the second of which is captured by rs849333 as the primary peak in the FHS. However, rs563289 is the secondary peak in FHS and appears to be a novel association, despite lying in the same physical region as the primary peak in CAGE.

Figure 4

Biological annotation of detected cis-eQTL signals. (A and B) Comparison of CADD score distributions for Primary and Secondary eSNPs and neighboring background SNPs at similar minor allele frequencies in the CAGE (A) and FHS (B) studies. (C and D) Relationship between observed beta and predicted deltaSVM score for significant peak eSNPs in the respective studies. P-values for comparisons are indicated.

Figure 5

Extent of replication of eQTL-GWAS colocalization with different expression platforms. The Venn diagram shows number of eQTL-GWAS joint associations (CLPP >0.001) in the three studies, and the percentage of all of the 3908 total associations in each sector.

Figure 6

Two examples of eQTL-GWAS colocalization. (A) Five loci with joint CAGE-eQTL and GWAS associations with CAD are joined in an extended network with seven additional loci in FHS. (B) IKZF3 was previously reported to be associated with three autoimmnune diseases; our analysis finds extended associations with other autoimmune diseases and blood traits. Genes are represented as solid dark green circles, and phenotypes or diseases as light blue circles. The thickness of edge line represents the strength of colocalization signal given by the indicated CLPP.