We further evaluated the type I error rate and power of GPA for the test of pleiotropy in our simulations. The simulation parameters were the same as those in the previous simulations. Power was evaluated at , , , and . The type I error rate was evaluated at , corresponding to the expected shared proportion of risk SNPs in the absence of pleiotropy. As shown in Figure 5, power increased as decreased and as and increased, whereas the type I error rate was appropriately controlled in all cases. Please note that type I errors and power remained almost the same for hypothesis testing of pleiotropy in the presence of annotation (see Figures S2–S4 in Text S1).
