Comparative study of three commonly used continuous deterministic methods for modeling gene regulation networks

BMC Bioinformatics

Table 4 Reverse-engineering results with data fitting scores only

Sparse data		ANN data	SS data	GRLOT data
to ANN model	P _ver	0.9925	0.9909	0.9926
	P _val	0.9939	0.9907	0.9900
	ΔP _fit	0.0014	-0.0002	-0.0026
to SS model	P _ver	0.9825	0.9986	0.9955
	P _val	0.6769	0.8822	0.6847
	ΔP _fit	-0.3056	-0.1164	-0.3108
to GRLOT model	P _ver	0.9737	0.9672	0.9975
	P _val	0.9712	0.9333	0.9970
	ΔP _fit	-0.0025	-0.0339	-0.0005
Detailed data
to ANN model	P _ver	0.9999	0.9846	0.9975
	P _val	0.9998	0.9789	0.9967
	ΔP _fit	-0.0001	-0.0057	-0.0008
to SS model	P _ver	0.9916	0.9999	0.9981
	P _val	0.8294	0.9917	0.8321
	ΔP _fit	-0.1622	-0.0082	-0.1660
to GRLOT model	P _ver	0.9936	0.9934	0.9986
	P _val	0.9935	0.9879	0.9982
	ΔP _fit	0.0001	-0.0055	-0.0004

Data (created by ANN, SS, GRLOT methods) predicted by the different methods (for cases A, B, C) and subsequently used for making in silico predictions; for sparse and detailed data sets.

ISSN: 1471-2105