Fig. 3

Case study 2: power under univariate and bivariate variance components models. This example shows the power to detect a single causal SNP using both univariate and bivariate variance components simulation models and the OpenMendel module for variance components analysis [8]. For each anlysis, each line in the graph depicts the power to detect a SNP with \(\text {MAF} = 0.23\) using 1000 simulations at significance level \(\alpha = 5 \times 10^{-8}\). The SNP effect size varies from 0 to 0.065 in increments of 0.002 in the center range (0.016–0.032) and increments of 0.005 in the two end ranges. On the x-axis, we convert the SNP MAF and effect sizes into the proportion of variation explained by the SNP. See the text for a detailed description of the model