Figure 4
From: Finding genes that influence quantitative traits with tree-based clustering
Power calculations. The left-hand plot gives a quantile-quantile plot for p-values from the analysis of data set 1. The solid line is the expected result. In the right-hand panel true-positive proportions are plotted against false-positives for varying significance level. This summarizes the results over all 200 replicate data sets, with solid lines denoting test statistics from five disjoint nodes (S5) and dotted lines denoting the maximum node value (S1).