Figure 2
From: Differential expression analysis with global network adjustment
Proof-of-concept simulated differential expression experiment. Histograms of p-values are shown for both differentially expressed genes (DE) and not DE genes that follow the null hypothesis. The raw (naive) model has p-values for not DE genes that do not have the expected uniform distribution, though the p-values corresponding to the differentially expressed (DE) genes tend to be low. The OSRR model applied to the same dataset more clearly separates the DE from the not DE genes, and the distribution of the not DE genes is more uniform.