NEW YORK – Genetic loci peppered over more than half of the genome are linked to complex human traits, with most influencing multiple traits, according to a new pleiotropy and genetic architecture analysis from an international team based in the Netherlands, Norway, and the US.
Despite the insights into genetic susceptibility, disease biology, and gene regulation that have been gleaned from GWAS done over the past decade and a half, the team explained, more complicated interactions between genetic risk variants, genes, and complex traits are less fully understood — from the number of traits impacted by a given gene or genetic variant to the genomic distribution of polygenic trait contributors.
"Such knowledge would greatly enhance our understanding of how genetic variation leads to trait variation and trait correlations," co-senior and corresponding author Danielle Posthuma, a complex trait genetics researcher affiliated with VU University in Amsterdam and the VU Medical Center, and her colleagues wrote in a paper published online today in Nature Genetics, noting that "the current availability of vast amounts of GWAS results allow investigation of these general questions."
For their own analyses, the researchers brought together summary statistics for nearly 4,200 genome-wide association studies in the searchable GWAS ATLAS database, including new genome-wide association analyses spanning 600 traits in tens of thousands of UK Biobank participants. By focusing on well-powered GWAS for 558 traits or conditions classified into two-dozen "domains," they analyzed pleiotropy, SNP features, and genetic architecture for complex trait-associated loci, identifying 41,533 of these loci across 60.1 percent of the genome.
More than 93 percent of loci showed ties to more than one trait, and 90 percent of the loci coincided with traits across multiple trait domains. Pleiotropy was particularly prominent at a stretch of chromosome 6 sequence containing the major histocompatibility complex, the team reported, which housed 441 trait-linked loci contributing to more than 200 traits.
When the researchers took a look at the SNPs involved with the complex trait associations — based on information across 1.7 million SNPs profiled for each of the 558 core trait GWAS — they saw significant associations involving almost 237,000 SNPs. Again, many SNPs had ties to multiple traits. Although 89 percent of the lead SNPs at trait-linked loci turned up in non-coding parts of the genome, the variants tended to come from regulatory regions and other sites near genes.
"We find that potential causal variants are enriched in coding and flanking regions, as well as in regulatory elements," the authors wrote, "and show variation in polygenicity and discoverability of traits."
At the gene level, at least 11,544 genes were associated with 518 of the traits considered, the researchers reported, including genes from a highly pleiotropic portion of chromosome 3. Genes known for being expressed in multiple tissue types also appeared more apt to pleiotropic associations with complex traits.
Among other analyses, the team considered the broader genomic architecture behind the associations documented across the GWAS and looked at the potential pleiotropy of sets of genes with shared functions, linking some 57 percent of these gene sets to multiple traits.
Based on the varying levels of pleiotropy they documented at the locus, gene, and SNP levels, the authors proposed that "a gene can be involved in multiple traits, but how that gene is affected by the causal SNPs may differ across traits."
"For instance," they explained, "the function of the gene can be disrupted through a coding SNP for one trait, while expression of that gene can be affected through a regulatory SNP for another trait. At the same time, overlap of trait-associated loci can be observed due to the overlap of the [linkage disequilibrium] blocks while each trait may be affected by the distinct genes."