Power of different methods in the simulation analysis.

Results were based on 200 replicates. In each replicate, 500 cases and 500 controls were used to identify three causal variants from a total of ∼1000 variants, with each method being evaluated. We assumed causal variants have a constant GRR of 3 and render disease susceptibility under a dominant model.1The proportion of replicate simulations in which at least one causal variant was detected.2The proportion of replicate simulations in which at least two causal variants were detected.3The proportion of replicate simulations in which all three causal variants were detected.4The average power of three causal variants.