STRUCTURE analysis using 174 AIMs with adequate quality excluded population stratification between cases and controls in this study (Fig. 2). It also shows clear distinction between the AA sample and the EA sample in our recent (Levran et al., 2008) (Fig. 3). The average extent of admixture (European ancestry proportion) in this AA sample population is 11%. Five AA individuals showed higher than expected EA admixture (0.64-0.84) and the analyses were repeated without them with no significant change in the results.
