Fig. 7

Power plot for the TCGA breast cancer data set. Power was calculated for both the LRT and Wald tests under the BNB model and the LRT under the Poisson-LMM model by using the empirical parametric test at μ=14,ϕ=1, and 11 different sample sizes (range 3-20). The nominal false positive rate α is set at 0.001. The detectable fold ratios of the mean expression levels between primary tumor and matched normal are set at 2 (solid lines) and 0.5 (dashed lines) under the alternative hypothesis