Manhattan plots of the association between SNPs and stable warfarin dose after pooling 6 African ancestry cohorts in a meta-analysis, gene-based analysis (n = 1504 participants). Individual GWAS analyses were undertaken using logarithm transformed stable warfarin dose, adjusted for age, sex, weight, target INR range, simvastatin/amiodarone status, and either the first 10 principal components of genetic ancestry or the proportion of the specific ancestry per chromosome by frequentist association testing assuming an additive model of inheritance before being pooled using METAL or MR-MEGA. The red horizontal lines represent the Bonferroni corrected significance thresholds (0.05 divided by the number of protein coding genes) and included 2.7 × 10−6 (18 244 genes), 2.7 × 10−6 (18 231 genes), and 2.7 × 10−6 (18 244 genes) for (A) METAL meta-analysis with standard GWAS, (B) METAL meta-analysis with GWAS using African ancestry tracts, and (C) MR-MEGA meta-analysis with standard GWAS, respectively. Some top genes are annotated. CYP2C9/18, cytochrome P450, family 2, subfamily C, polypeptide 9/18; KAT8, lysine acetyltransferase 8; VKORC1, vitamin K epoxide reductase complex, subunit 1.
