Table II.

Performance of the BNP model on three control data sets

Data seta and methodb	Expected FDR = 5%			Expected FDR = 1%
	Actual FPR	Total/correct	Sensitivity	Actual FPR	Total/correct	Sensitivity
	%		%	%		%
D1
M1	2.23	719/703	78.29	0.53	567/564	62.81
M2	2.59	733/714	79.51	0.89	674/668	74.39
M3	2.20	820/802	89.31	0.40	758/755	84.08
M4	2.72	810/788	87.75	1.39	722/712	79.29
D2
M1	1.92	5,875/5,762	68.20	0.36	4,964/4,946	58.54
M2	2.17	6,775/6,628	78.45	0.51	5,895/5,865	69.42
M3	3.16	7,426/7,191	85.11	1.04	6,754/6,684	79.11
M4	1.91	7,001/6,867	81.28	0.55	6,333/6,298	74.54
D3
M1	0.13	10,284/10,271	74.80	0.03	9,182/9,179	83.70
M2	0.42	11,477/11,429	93.14	0.17	10,699/10,681	87.04
M3	0.50	11,983/11,923	97.16	0.09	11,388/11,378	92.72
M4	0.32	10,885/10,850	88.42	0.16	10,117/10,101	82.32

^aD1, LCQ control data set; D2, LTQ control data set; D3, LTQ/FT control data set. For details see “Experimental Data Sets.”

^bM1, cutoff-based method; M2, PeptideProphet; M3, BNP model; M4, nonparametric model.