Table 1.

Parameters used to run the simulations of the power analyses. All possible combinations of these parameters (60 total) were run.

Coverage (X)	Coefficient of variation (CV)c	Number of replicatesd	Effect size (log-fold change)e	Alfa (threshold p-value)
20a	0.2	3	1.5	0.05
10b	0.3	4	1.75
	0.4		2
			3
			4

^abased on the Lander/Waterman estimate of coverage for our data; i.e., mean coverage of 23X per replicate (range 19–28 among our replicates).

^bbased on our minimum number of reads per replicate (in one treatment) required for a DE test (see text)

^cThese values encompass the range of most of the values of Biological Coefficient of Variation (BCV; estimated by edgeR) obtained (i.e., a “Common” value of ~0.31 and a stabilized “Trend” value of ~0.2; see Fig. 2).

^dreflects the range of replicates per treatment used in this study

^epower was high and stable beyond an effect size of 4 (see Fig. 5). Thus, higher effect sizes are not reported.