BH vs Permutations: False Discovery Rate Estimation Comparison #1937229 (License: Personal Use)

This plot evaluates two multiple testing correction strategies-Benjamini-Hochberg (BH) and permutation-based FDR estimation-by plotting estimated FDR against true FDR. The dashed diagonal represents perfect calibration; both methods lie below it, indicating conservative estimates, with permutations slightly less conservative than BH across most of the range. Each point reflects aggregated simulation results, with confidence bands implied by the shaded regions.

Used in statistical methodology papers, bioinformatics tutorials, or genomics analysis guides to illustrate FDR control performance. Targets researchers evaluating correction methods for high-dimensional data like RNA-seq or GWAS.