Sample Size Calculation for Controlling False Discovery Proportion

Table 4

Sample size calculation using the prostate cancer dataset.

𝜋0	𝑚1	𝑛	P̂(FDP≤0.10)	P̂((𝑈/𝑚1)≥0.9)	𝛼̂
0.95	500	74	0.830	0.915	0.003
0.9	1000	63	0.855	0.970	0.007
0.85	1500	57	0.885	0.955	0.012
0.8	2000	52	0.875	0.945	0.018
0.75	2500	48	0.870	0.955	0.026
0.7	3000	44	0.915	0.925	0.034

P̂(FDP≤0.10) and P̂((𝑈/𝑚₁)≥0.9): empirical probability from 200 simulation runs.