We performed principal component analysis (PCA) to assess allele frequency differences between cases and controls due to ancestry differences and statistical analyses were performed using EIGENSTRAT [29], [30]. This software detects population structure inferring axes of genetic variation and outputs each individual's coordinates along axes of variation. The analysis was performed using a panel of 64 SNPs ancestry-informative markers (AIMs) and models were created with two to five principal components in order to detect the existence or absence of population structure; details of marker set are available on request. In our data set, we were not able to demonstrate a difference in population stratification between cases and controls therefore, no corrections using the PCA results were done in the association tests for the BCHE markers (Figure 1).
