Genome-wide association studies (GWAS)
GWAS were conducted using logistic regression (LR). PC scores were estimated with PLINK and used to adjust the population substructure. A total of 20 PC scores corresponding to the 20 largest eigenvalues, age and sex were included as covariates. The genome-wide significance level was set to 5×10-8. Genome-wide significant SNPs were annotated with ANNOVAR. We used LDlink19 to evaluate the linkage disequilibrium (LD) across the ancestral population with Phase 3 haplotype data from the 1000 Genomes Project.