Primo: integration of multiple GWAS and omics QTL summary statistics for elucidation of molecular mechanisms of trait-associated SNPs and detection of pleiotropy in complex traits

Abstract

To provide a comprehensive mechanistic interpretation of how SNPs affect complex traits, we propose a method -Primo -for integrative analysis of GWAS summary statistics with multiple sets of omics QTL summary statistics from different cellular conditions or studies. Primo identifies SNPs in various association patterns to complex and omics traits and performs conditional association analysis in a region to account for linkage disequilibrium. Primo allows for unknown study heterogeneity and sample correlations. We show two types of applications using Primo to examine the molecular mechanisms of known susceptibility loci and to detect and interpret pleiotropic effects.

CCmed: cross-condition mediation analysis for identifying robust trans-eQTLs and assessing their effects on human traits

“…To quantify the cross-tissue cis-association probability, Pr(α c = 0 in all K tissue types) in (1), we first obtained the gene-level cis-association statistics for M cis-genes by F -tests. We applied a recently developed integrative association analysis approach, Primo [15,21], to the association statistics to obtain the probabilities of Pr(α c = 0 in all K tissue types) for gene i = 1, . .…”

“…For each (GWAS SNP, cis-gene) pair, we conducted regression (6) described in the Methods, regressing cis-gene expression on the GWAS SNP genotype and adjusting for GTEx-reported eQTLs, to obtain the cis-association statistics in each tissue type. We applied Primo [15] to the M g ×K matrix of cis-association statistics {F G } where M g is the number of (GWAS SNP, cis-gene) pairs, and estimated the probability of each GWAS SNP being also an eQTL in at least 2 tissue types, conditioning on other cis-eQTLs. There were 40 (GWAS SNP, cis-gene) pairs showing evidence of cis-association in at least 2 tissue types (with cross-2-tissue cis-association posterior probability > 80%).…”

“…For each of the 40 (GWAS SNP, cis-gene) pairs, we further conducted conditional correlation analysis with each of its trans-genes in each tissue type using the regression model (7) described in Methods to obtain a matrix of conditional correlation statistics, {Z G i.. }, conditioning on cis-eQTL and GWAS SNP genotypes. We applied Primo [15] to {Z G i.. } to estimate the probability of non-zero conditional cis-trans gene-expression correlation for each possible pair of tissues s . Using equation (2), we obtained a lower bound estimate of the probability of cis-mediated trans-association in at least two brain tissue types for each trio, and identified 1492 (GWAS SNP, cis-gene, trans-gene) trios with P GWAS med,ij > 80% (FDR = 7.9%) with 1418 unique trans-genes.…”

“…To estimate Pr(α c = 0 in all K tissues), we apply a recently developed integrative association analysis approach, Primo [15,21]. Primo takes as input the M × K matrix of cis-association summary statistics {F ik }, considers all 2 K possible crosstissue association patterns and estimates the density function for each pattern, and returns the probabilities of P 1i = Pr(α c = 0 in all K tissue types) for gene i = 1, .…”

“…In addition to mapping robust gene-level trans-associations in the genome, another major focus of this work is to study the trans-genes of GWAS SNPs and further understand how genetic variation affects complex traits -a long-standing question in genetics research. Increasing evidence has suggested that many susceptibility loci may affect complex traits/diseases via the modulation of their cisgene expression levels [14,15,16,17]. Those cis-gene expression levels may further affect other distal genes in functional pathways, some of which also con-tribute to phenotypic variation.…”