Table 1. Number of gene-chip predicted SFPs called at different threshold (delta) in datasets.

(a) CP&LG					(b) CP&RT					(c) LG&RT
Delta	p0	FALSE	Called	FDR	Delta	p0	FALSE	Called	FDR	Delta	p0	FALSE	Called	FDR
1	0.95	3128.7	13668	0.217	1	0.95	16942.4	110346	0.146	1	0.95	12349	92926	0.126
2	0.95	1134.4	9093	0.119	2	0.95	6457.15	61055	0.100	1.2	0.95	9652.9	80935	0.113
3	0.95	719.8	6865	0.100	3	0.95	4558.5	46306	0.094	1.4	0.95	8131.8	72951	0.106
4	0.95	537.7	5376	0.095	4	0.95	3580.15	37137	0.092	1.6	0.95	7165.1	67299	0.101
5	0.95	423.55	4274	0.094	5	0.95	2905.7	30322	0.091	1.8	0.95	6493.4	62756	0.098
6	0.95	344.85	3488	0.094	6	0.95	2404.95	25325	0.090	2	0.95	5983.9	58918	0.096

The data were analyzed as described [7]. (p0 = the prior probability of the proportion of SFP in the null datasets; Called = the number of SFP at each threshold; False = the number of SFP in the mean permuted dataset; FDR = false discovery rate; CP = Cypress; LG = LaGrue and RT = RT0034.