Table 1.
Mean type-I error of the methods.
Table 2.
Mean power of the methods for the different pathway sizes.
Figure 1.
Power of the five methods for different pathway sizes.
Plots illustrate the power of the methods when the total number of associated SNPs equals 100, the proportion of associated SNPs within each pathway is 0.4 and the effect sizes are and
.
Table 3.
Power of the methods as the proportion of pathway SNPs with effects changes.
Figure 2.
Power of the five methods for different proportions () of associated SNPs within a pathway of size 60.
Plots illustrate the power of the methods when the total number of associated SNPs equals 100 (plot (A)) and 200 (plot(B)), the effect sizes are and
for both plots.
Table 4.
Power of the methods for the different pathway sizes.
Table 5.
Mean power of the methods that have a type-I error 5% across all simulated scenarios.
Figure 3.
Pairwise scatterplot of power for the five methods across all simulated non-null scenarios.
Table 6.
Pathway Analysis of BMI.
Table 7.
Performance of the methods when applied on the data of the GWAS.