Skip to main content
. 2015 May 12;44(4):1429–1441. doi: 10.1093/ije/dyv041

Figure 5.

Figure 5.

DNA methylation variance and correlation can impact EWAS power. Case-control power estimates (A, upper panel) are shown under different pooled SDs and methORs at a fixed mean difference = 10% using parametric (left panel) and nonparametric (right panel) test statistics. MZ twin power estimates (B, lower panel) are shown under different pooled SDs and correlation coefficients at a fixed mean difference = 9% using parametric (left panel) and nonparametric (right panel) test statistics.