This example compares SNP-based PCA to QGS-based PCA. It uses public HapMap3 data to demonstrate that the first two principal components from SNPs and QGS are close to identical (a correlation of r>0.99). In addition, it creates two population stratification plots based on the two PCs. This analysis is used in the QGS publication.
The Linux shell script hm3.sh downloads and extracts the HapMap3 data, converts the data to Plink binary format using the plink command line tool, calculates QGS values, and calls R to do the PCA.
The R script hm3.R calculates PCs for both the SNPs and QGS and plots these.
The two pdf files contain the output of the script: two population stratification plots.