Properties of local interactions and their potential value in complementing genome-wide association studies

Abstract

Local interactions between neighbouring SNPs are hypothesized to be able to capture variants missing from genome-wide association studies (GWAS) via haplotype effects but have not been thoroughly explored. We have used a new high-throughput analysis tool to probe this underexplored area through full pair-wise genome scans and conventional GWAS in diastolic and systolic blood pressure and six metabolic traits in the Northern Finland Birth Cohort 1966 (NFBC1966) and the Atherosclerosis Risk in Communities study cohort (ARIC). Genome-wide significant interactions were detected in ARIC for systolic blood pressure between PLEKHA7 (a known GWAS locus for blood pressure) and GPR180 (which plays a role in vascular remodelling), and also for triglycerides as local interactions within the 11q23.3 region (replicated significantly in NFBC1966), which notably harbours several loci (BUD13, ZNF259 and APOA5) contributing to triglyceride levels. Tests of the local interactions within the 11q23.3 region conditional on the top GWAS signal suggested the presence of two independent functional variants, each with supportive evidence for their roles in gene regulation. Local interactions captured 9 additional GWAS loci identified in this study (3 significantly replicated) and 73 from previous GWAS (24 in the eight traits and 49 in related traits). We conclude that the detection of local interactions requires adequate SNP coverage of the genome and that such interactions are only likely to be detectable between SNPs in low linkage disequilibrium. Analysing local interactions is a potentially valuable complement to GWAS and can provide new insights into the biology underlying variation in complex traits.

Figure 1. Differences in the numbers of SNP pairs with P_int<5.0E-08 and local interaction pairs (P_int<1.0E-05) detected in each trait between ARIC and NFBC1966.

Figure 2. A cartoon model illustrating a haplotype tagging a recessive causal variant can generate an apparent statistical interaction between two unlinked SNPs each with limited marginal effects.

(I) A recessive causative variant (black star) is associated with only the ab SNP haplotype, assuming Hardy-Weinberg Equilibrium, i.e. an equal allele frequency of 0.5 for each SNP so there is no LD between the two SNPs and an equal frequency of 0.25 for each of the four possible haplotypes, and the causal variant with an effect size of 1. (II) Only individuals homozygous for this haplotype (ab/ab) are genetically differentiated generating apparent epistasis (averaged trait value and joint genotype frequency in the bracket in each cell). (III) Marginal effects associated with the individual SNPs are limited with only one in four individuals of the aa or bb SNP genotype being affected with a trait value of 2 so the averaged trait value of the genotype is 0.5 (SNP genotype frequency in brackets), thus the individual SNPs may not be detected by a conventional GWAS. (IV) This resembles the interaction between rs17119975 and rs10892020 in TRI (Table 1) where neither SNP had important marginal effects and their interaction signal was mainly because of the differentiated phenotype associated with the double homozygous aabb genotype.

Figure 3. Proportions of local interactions in different LD ranges.

Figure 4. Local interactions (P_int<1.0E-05) within the 11q23.3 region associated with TRI in ARIC and supporting ENCODE regulatory evidence.

(A) black oval: the top marginal SNP rs964184 within the region; each line representing an interaction between two SNPs at the start and end locations where red and blue lines represent interactions prior to and post conditional tests respectively; red and blue ovals: the marginal SNP rs6589567 prior to and post conditional test respectively; y axis: association P values in the −log₁₀ scale; x axis: genomic location in base pair; arrow bar showing transcription direction and location of the gene (italic) below the bar. (B) a snapshot from UCSC genome browser showing clustered ENCODE regulatory elements aligning to the 5′ ends of BUD13 and ZNF259 respectively.