Boosting the power of genome-wide association studies within and across ancestries by using polygenic scores
Journal
Nature Genetics
Journal Volume
55
Journal Issue
10
Pages
1169
Date Issued
2023-10
Author(s)
Campos, Adrian I
Namba, Shinichi
Lin, Shu-Chin
Nam, Kisung
Sidorenko, Julia
Wang, Huanwei
Kamatani, Yoichiro
Wang, Ling-Hua
Lee, Seunggeun
Lin, Yen-Feng
Okada, Yukinori
Visscher, Peter M
Yengo, Loic
Abstract
Genome-wide association studies (GWASs) have been mostly conducted in populations of European ancestry, which currently limits the transferability of their findings to other populations. Here, we show, through theory, simulations and applications to real data, that adjustment of GWAS analyses for polygenic scores (PGSs) increases the statistical power for discovery across all ancestries. We applied this method to analyze seven traits available in three large biobanks with participants of East Asian ancestry (n = 340,000 in total) and report 139 additional associations across traits. We also present a two-stage meta-analysis strategy whereby, in contributing cohorts, a PGS-adjusted GWAS is rerun using PGSs derived from a first round of a standard meta-analysis. On average, across traits, this approach yields a 1.26-fold increase in the number of detected associations (range 1.07- to 1.76-fold increase). Altogether, our study demonstrates the value of using PGSs to increase the power of GWASs in underrepresented populations and promotes such an analytical strategy for future GWAS meta-analyses.
SDGs
Type
journal article