Imputation is a statistical method used to determine probabilistically missing or untyped genotypes using a densely mapped reference panel (26–28). We performed imputation-based analysis in our ethnic-specific case-control samples using publicly available 1000Genomes as data reference panels (2010–11 1000G Interim Phase I) for imputation, including 246 AA (YRI+LWK+ASW), 381 EA (GBR+FIN+IBS+CEU+TSI), 181 HS (PUR+CLM+MXL), and 286 AS (JPT+CHB+CHS). For HapMap family data only founders were included in the reference panels. From this region (114,356,663–114,418,205 bp) 527 SNPs were available. Imputation was performed using MACH [42], which provided a quantitative assessment of estimate uncertainty (Rsq). Allelic association results were adjusting for imputation uncertainty with mach2dat. All imputed SNPs were filtered based on QC measures (HWE >0.001, MAF >0.01, Rsq >0.6).
