Fig. 2From: Integrating mean and variance heterogeneities to identify differentially expressed genesComparison of false positive rates of eight methods under standard normality setting. Each panel was computed from 100,000 replicates of two-group samples with the specified samples sizes simulated from \( \mathcal{N}\left(0,1\right) \). At each significance level, the false positive rate of each method was estimated by the empirical proportion that the method rejected the dual null hypothesis H03. The gray belt is the 95% concentration band of the false positive rates of a typical test that can properly control false positive rates at given nominal significance levels. a n1=n2=5, b n1=n2=10, c n1=n2=20, d n1=n2=40Back to article page