Table 2.

Sample size and anticipated average power calculated by our method and observed average power by the voom and limma pipeline while controlling FDR using q-value procedure based on parameters at different m for three simulation settings

	Simulation 1			Simulation 2					Simulation 3
	Our Method			Our Method			RnaSeqSampleSize		Our Method
m	Sample	Anticipated	Observed	Sample	Anticipated	Observed	Sample	Estimated	Sample	Anticipated	Observed
	size	power	power	size	power	power	size	power	size	power	power
50	35	0.802	0.858	13	0.810	0.876	9	0.780	22	0.800	0.801
100	34	0.814	0.846	13	0.814	0.876	9	0.724	22	0.803	0.801
200	34	0.815	0.846	13	0.823	0.876	9	0.765	22	0.804	0.801
500	33	0.817	0.833	13	0.827	0.876	9	0.769	22	0.805	0.801
1000	32	0.800	0.804	13	0.826	0.876	9	0.764	22	0.805	0.801

We also present the comparison between our method and RnaSeqSampleSize R package for simulation 2, where the right two columns are sample size and power calculated by the RnaSeqSampleSize R package