Table 5.

Number of significant p-values (tests) found in the “M” vs. “S” treatment contrast of the real proteomics dataset as a function of test-statistics, FDR control procedure, and scale (RAW or LOG) at a fixed α level of False Discovery Rate (FDR ≤ α, α ∈ {0.01, 0.05}). Results are reported for each procedure with either Benjamini-Yekutieli's FDR control procedure under variable dependency (BY), Benjamini-Hochberg's regular FDR control procedure (BH), or John Storey's robust computation of the positive pFDR (JS). For the computation of the p-values, we followed the resampling scheme of Subsection 4.3 with B′ = 256 bootstrap samples.

	t – REG						t – MVR
	BY		BH		JS		BY		BH		JS
	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05
LOG	0	0	0	0	0	0	11	11	11	11	11	9052
RAW	0	0	0	0	0	0	7	7	7	7	5611	9052

	t – CVSR						t – HBR
	BY		BH		JS		BY		BH		JS
	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05
LOG	0	0	0	0	0	0	5	17	23	75	23	78
RAW	6	6	6	6	6	6	0	0	0	0	0	0

	t – EBR						t – SR
	BY		BH		JS		BY		BH		JS
	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05
LOG	0	0	0	0	0	0	0	0	0	0	0	0
RAW	3	3	3	3	3	3	0	0	0	0	0	0

	t – BR						t – JSR
	BY		BH		JS		BY		BH		JS
	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05
LOG	1	1	7	52	7	52	0	0	1	1	1	1
RAW	0	0	0	0	0	0	0	0	0	0	0	0

	t – VSN						t – LOESS
	BY		BH		JS		BY		BH		JS
	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05
LOG	0	0	0	0	0	0	0	0	0	0	0	0
RAW	0	0	0	0	0	0	0	0	0	0	0	0

	t – CSS						t – QUANT
	BY		BH		JS		BY		BH		JS
	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05	0.01	0.05
LOG	0	0	0	0	0	0	0	0	0	0	0	0
RAW	0	0	0	0	0	0	0	0	0	0	0	0

	t – GAMLSS
	BY		BH		JS
	0.01	0.05	0.01	0.05	0.01	0.05
LOG	0	0	0	0	0	0
RAW	0	0	0	0	0	0