GWAS on short tandem repeats identifies genetic mechanisms in Alzheimer’s disease

GWAS typically focus on SNPs, often excluding complex genetic variants, such as short tandem repeats. Here, we report the results of GWAS analyses systematically assessing the role of short tandem repeats, both imputed and directly genotyped by whole genome sequencing, on risk for Alzheimer's disease in a large collection of ~330,000 individuals (3287 cases; 47,048 Alzheimer's disease-by-proxy cases, 283,111 controls) from the UK biobank. Using short tandem repeat genotype data, we identify 15 independent loci showing evidence for genome-wide significant association with Alzheimer's disease risk. While most identified loci had already been highlighted by SNP-based GWAS, we detect short tandem repeat-based signals near the genes SNX32 (chr. 11q13) and WSB1 (chr. 17q11). In addition, we delineate several other loci where short tandem repeats (and not SNPs) either represent the lead signal (ABCA7) or make substantial contributions to the SNP-driven associations (HLA-DRB1, MINDY/ADAM10, and APOE). Heritability analyses estimate that short tandem repeats account for at least 3% of the total phenotypic variance of Alzheimer's disease in this dataset. Aligning our top short tandem repeats with DNA methylation and transcriptome profiles from human brain samples suggests that several short tandem repeats may unfold their effects by impacting gene expression.