Skip to main content
Premium Trial:

Request an Annual Quote

GWAS Using Million Veteran Program Data Examines Genetic Architecture of Thousands of Traits

NEW YORK – In a series of genome-wide association studies using data on hundreds of thousands of diverse participants from the Department of Veterans Affairs' Million Veteran Program (MVP), a University of Pennsylvania- and VA-led team has uncovered a large degree of similarity in the genetic architecture of thousands of traits across different ancestry groups. At the same time, they noted that some causal variants they identified appear to be population specific.

"Findings from this study aim to expand the current knowledge base on population genetic architecture and to underscore the importance of diversity in genetic research in uncovering the full spectrum of human genetic variation and its impact on complex traits," the authors explained in a study published in Science on Thursday.

For their analyses, researchers from the University of Pennsylvania-Perelman School of Medicine, the Corporal Michael J. Crescenz VA Medical Center, and elsewhere brought together genotyping data for 635,969 MVP participants, using individuals' genetic similarity to global populations to estimate ancestry and ties to groups in different parts of the world.

"The findings from this study, one of the largest and most diverse of its kind, will allow researchers to uncover critical insights that will drive better health outcomes for veterans from all backgrounds," coauthor and Million Veteran Program Director Sumitra Muralidhar said in an emailed statement, adding that the work "helps pave the way for advances in how we deliver the personalized healthcare our diverse Veteran population deserves."

Muralidhar noted that findings from the study "will also benefit Veteran families, caregivers, survivors, and other non-veterans."

In particular, the team estimated that the participant group included 449,042 individuals of European ancestry; 121,177 African ancestry individuals; 59,048 admixed American individuals; and 6,702 participants of East Asian ancestry.

Even so, the authors noted that while the classifications "are not intended to be deterministic of race or ethnic identities, they are inextricably intertwined with these social constructs."

"Our intent in applying categorical population descriptors is to facilitate the study of genetic variation and its association with traits and diseases between diverse populations," they explained, noting that the groupings "oversimplify the rich and complex mosaic of human genetic diversity."

With the help of population-specific genome-wide association studies and GWAS meta-analyses, which were informed by data across some 42 million SNPs and spanned some 1,847 phenotype codes, half a dozen vital sign measurements, 63 types of laboratory measures, and results from 240 survey questions, the team focused in on more than 26,000 variant associations involving 1,270 traits and some 13,672 genomic risk loci.

"Our comprehensive phenome-wide GWAS presented here underscores the increased power to discovery that comes from including individuals from diverse populations," the authors wrote, "enriching our understanding of the genetics of complex health and disease traits, while highlighting the large degree of similarities in genetic architecture of these traits across populations."

More than 1,600 risk loci appeared to be specific to non-European populations, the researchers reported. Likewise, when they used the SuSiE ("sum of single effects") model- and linkage disequilibrium-based fine-mapping to narrow in on 6,318 causal variants with ties to more than 600 traits, a subset of some 2,069 causal variants turned up specifically in individuals from non-European populations.

"Our analyses of variant-trait associations detected several signals that would not have been identified in a GWAS comprising solely individuals genetically similar to [European] reference populations," the authors reported, adding that findings from the study "exemplify how the inclusion of individuals from diverse populations in human genetics experiments generates important insights into health and disease traits that may disproportionately affect these groups."

In the African population-related individuals, for example, the team tracked down a SNP at the SLC22A18/SLC22A18AS locus that was linked to keloid scarring — an association that was not picked up in other populations.

Nevertheless, the researchers emphasized that genetic associations tended to turn up across distinct population groups, with variable allele frequency or allele representation differences accounting for most of the between-population differences detected.

Together, the authors reported, the new results "reveal more similarities than differences in genetic architecture across populations, with most differences attributable to allele frequency variations between populations."

"By making this information available to the global health care community, we can empower an effort to study conditions that impact Veterans across the spectrum of ethnic heritage and innovate new or better medical treatments for minorities," Carolyn Clancy, the assistant under secretary for health for Discovery, Education, and Affiliate Networks at the VA, said in an email.

In a related commentary in Science, Queen Mary University of London researcher Alice Williamson and Segun Fatumo, with the Queen Mary University of London and Uganda Research Unit, who were not involved in the study, agreed that the work offers a more complete look at genetic risk contributors in individuals of non-European ancestry.

Though Williamson and Fatumo cautioned that further studies are needed to look at the genetics of conditions prevalent in younger and female individuals, who tend to be underrepresented in the MVP, they called findings from the study "a valuable complement to other large-scale biobank efforts."