Skip to main content
. 2013 Apr 11;74(3-4):165–171. doi: 10.1159/000346022

Table 4.

Power comparisons for the different methods under the various effect and sample sizes considered when there exist both protective and risk variants

KATE
GRANVIL SKAT
weights = 1 weights ~ beta
N = 1,000
|beta|=1.2 0.625 0.665 0.220 0.999
|beta|=2.0 0.921 0.934 0.367 1
|beta|=3.0 0.969 0.973 0.480 1
|beta| ∝ MAF 0.331 0.393 0.147 0.970

N = 2,000
|beta|=1.2 0.891 0.909 0.380 0.999
|beta|=2.0 0.982 0.983 0.505 1
|beta|=3.0 0.991 0.994 0.621 1
|beta| ∝ MAF 0.577 0.636 0.213 0.998

N = 3,000
|beta|=1.2 0.952 0.960 0.426 1
|beta|=2.0 0.987 0.990 0.599 1
|beta|=3.0 0.988 0.991 0.672 1
|beta| ∝ MAF 0.769 0.816 0.292 1