Skip to main content

Advertisement

Table 3 Power and type I error assessment

From: Risk score modeling of multiple gene to gene interactions using aggregated-multifactor dimensionality reduction

A) Only Locus1*2 has an interaction Model 1 (n=300) Model 2 (n=300) Model 3 (n=300) Model 4 (n=300)
Power Type I Power Type I Power Type I Power Type I
A-MDR pOR 1 0.01 1 0.01 0.75 0.01 1 0.03
A-MDR pRR 1 0 1 0.01 0.71 0.01 1 0.03
A-MDR pChi 1 0.01 1 0 0.86 0.03 1 0
MDR 1 0 1 0 0.46 0.02 1 0
A) Only Locus1*2 has an interaction Model 1 (n=400) Model 2 (n=400) Model 3 (n=400) Model 4 (n=400)
Power Type I Power Type I Power Type I Power Type I
A-MDR pOR 1 0 1 0 0.86 0 1 0.02
A-MDR pRR 1 0 1 0 0.87 0 1 0.03
A-MDR pChi 1 0.02 1 0.01 0.95 0.01 1 0.02
MDR 1 0 1 0 0.68 0.01 1 0
B) Genetic heterogeneity γ 1 = γ 2 =0.5 Model 1 Model 2 Model 1 Model 2 Model 3 Model 4 Model 3 Model 4
n=400 Power n=800 Power n=400 Power n=800 Power
A-MDR pOR 0.98 1 1 1 0.37 0.91 0.7 1
A-MDR pRR 0.99 1 1 1 0.35 0.78 0.68 1
A-MDR pChi 0.99 1 1 1 0.31 0.99 0.72 1
MDR 0.12 0.75 0.08 0.78 0 0.69 0 0.92
B) Genetic heterogeneity γ 1 =0.7,γ 2 =0.3 Model 1 Model 2 Model 1 Model 2 Model 3 Model 4 Model 3 Model 4
n=400 Power n=800 Power n=400 Power n=800 Power
A-MDR pOR 1 0.35 1 0.98 0.55 0.35 0.92 0.78
A-MDR pRR 1 0.25 1 0.97 0.55 0.3 0.91 0.74
A-MDR pChi 1 0.5 1 0.99 0.6 0.49 0.96 0.86
MDR 1 0 0.99 0 0.14 0.21 0.14 0.38
C) Additive Models Model 1 Model 2 Model 1 Model 2 Model 3 Model 4 Model 3 Model 4
n=400 Power n=800 Power n=400 Power n=800 Power
A-MDR pOR 1 0.72 1 1 0.6 0.16 0.93 0.61
A-MDR pRR 1 0.69 1 0.98 0.61 0.14 0.91 0.55
A-MDR pChi 1 0.79 1 1 0.71 0.29 0.98 0.76
MDR 0.91 0.02 0.99 0.01 0.26 0.14 0.43 0.12
  1. (Within each cell, loci 1x2 is listed in the first column and loci 4x5 is in the second column).