As technology develops and genotyping chips become denser it is a natural question to ask how much power would be gained by genotyping additional SNPs or by using genotyping imputation methods [14]. To facilitate such comparisons we evaluated the performance of a hypothetical chip that contains all the SNPs in HapMap to act as a point of reference in our results. The performance of this ‘complete’ chip is shown as a solid black line in all of the figures showing power. Since the simulations we carry out only use HapMap SNPs as causal SNPs this analysis approximates the scenario in which we have a chip which types all possible SNP variation.
