Power of Matched versus Unmatched PC Adjusted GWAS. We compared the power of a matched GWAS (2000 cases/2000 controls) to an unmatched PC adjusted GWAS with a varying number of controls (2000 to 4000). The x-axis reports the number of controls added to the unmatched control group (n = 2000) and the letter M corresponds to the matched GWAS. Power was assessed for a MAF ranging between 0.1 to 0.5 and a relative risk (RR) ranging from 1.2 to 1.5. In all scenarios the matched analysis is more powerful than the unmatched analysis when the sample is the same for both analyses. As we add more controls to the unmatched analysis, the power increases and surpasses the power of the matched analysis when 1000 to 1500 additional controls are added. Note that the power is adjusted for the varying false positive rate across scenarios.
Solovieff et al. BMC Genetics 2010 11:108 doi:10.1186/1471-2156-11-108