RSNET: inferring gene regulatory networks by a redundancy silencing and network enhancement technique

BMC Bioinformatics

Table 2 The results on DREAM networks with scales 10, 50 and 100

Approach	FPR	TPR	ACC	PPV	MCC	AUC
Scale 10
LASSO	0.8375	0.6000	0.2111	0.0822	− 0.1907	0.7025
LP	0.4125	0.1000	0.5333	0.0294	− 0.2026	0.7388
RO	0.5000	0.1000	0.4556	0.0244	− 0.2524	0.7975
GENIE3	0.1125	0.7000	0.8889	0.5000	0.6187	0.9212
ARACNE	0.1125	0.9000	0.8889	0.5000	0.6187	0.9300
NARROMI	0.1125	0.9000	0.8889	0.5000	0.6187	0.9294
RSNET	0.0375	0.7000	0.9333	0.7000	0.6625	0.9450
Scale 50
LASSO	0.1285	0.3506	0.8551	0.0813	0.1132	0.7110
LP	0.0847	0.3896	0.8988	0.1299	0.1820	0.6686
RO	0.1311	0.4935	0.8571	0.1089	0.1809	0.7268
GENIE3	0.0745	0.4805	0.9114	0.1729	0.2508	0.8004
ARACNE	0.0817	0.5974	0.9082	0.1917	0.3027	0.8325
NARROMI	0.0623	0.5325	0.9249	0.2169	0.3074	0.8389
RSNET	0.0594	0.5195	0.9273	0.2210	0.3069	0.8376
Scale 100
LASSO	0.0510	0.1807	0.9361	0.0569	0.0741	0.6536
LP	0.0462	0.4036	0.9445	0.1296	0.2063	0.6741
RO	0.0854	0.3735	0.9055	0.0693	0.1290	0.6856
GENIE3	0.0255	0.4096	0.9657	0.2315	0.3076	0.8407
ARACNE	0.0330	0.5060	0.9592	0.2069	0.3062	0.8572
NARROMI	0.0243	0.4639	0.9671	0.2452	0.3220	0.8584
RSNET	0.0259	0.5120	0.9663	0.2515	0.3437	0.8594

ISSN: 1471-2105