Local Ancestry Corrects for Population Structure in Saccharomyces cerevisiae Genome-Wide Association Studies

Supporting Information for Diao and Chen, 2012

Supporting Information

• Supporting Information - Figures S1-S6 and Table S1 (PDF, 1.6 MB)
• Figure S1 - This figure shows a comparison of representative runs of the program, STRUCTURE, using SNPs selected according to the windowed (top) and sequential (bottom) methods (PDF, 251 KB)
• Figure S2 - This figure shows that SNPs selected by the windowing method (bottom) tend to be more uniform across the genome whereas SNPs selected by the sequential method (top) tend to follow the recombination landscape of the genome more closely (PDF, 218 KB)
• Figure S3 - Each S. cerevisiae strain is plotted on the first two principal components, after performing principal components analysis on the genotype matrix (PDF, 228 KB)
• Figure S4 - ROC curve of simulations with 100 causal SNPs each adding phenotypic value 1 (PDF, 393 KB)
• Figure S5 - ROC curve of simulations with 3 causal SNPs each adding phenotypic value 10 (PDF, 328 KB)
• Figure S6 - ROC curve of simulations with 3 causal SNPs, each adding phenotypic value 10, with a base phenotypic value given by one trait from (WARRINGER et al. 2011) (PDF, 359 KB)
• Table S1 - This table gives summary statistics of the mean squared distance (MSDs) of all GWAS methods used over all phenotypes (PDF, 43 KB)