Table 2.
Power Comparison with Existing Pathway Approaches
Set | G1 | p2 | Causal SNPs | α Level |
Power |
|||
---|---|---|---|---|---|---|---|---|
Wang | Plink | ALIGATOR | GRASS | |||||
E1 | 20 | 444 | 53 | 0.05 | 0.48 | 0.68 | 0.41 | 0.87 |
0.01 | 0.24 | 0.35 | 0.24 | 0.79 | ||||
E2 | 20 | 444 | 54 | 0.05 | 0.54 | 0.92 | 0.55 | 0.91 |
0.01 | 0.18 | 0.50 | 0.16 | 0.80 | ||||
F1 | 30 | 2286 | same as in E1 | 0.05 | 0.29 | 0.44 | 0.18 | 0.87 |
0.01 | 0.20 | 0.23 | 0.02 | 0.80 | ||||
F2 | 30 | 2286 | same as in E2 | 0.05 | 0.32 | 0.80 | 0.33 | 0.87 |
0.01 | 0.17 | 0.61 | 0.13 | 0.73 |
Summary of power comparison of Wang et al.,7 Plink,18 ALIGATOR,9 and the proposed GRASS under various simulated scenarios.
Number of genes.
Total number of SNPs.
One causal SNP in each of the five smallest genes (range: 4–11 SNPs).
One causal SNP in each of the five largest genes (range: 28–91 SNPs).