PFBNet: a priori-fused boosting method for gene regulatory network inference

BMC Bioinformatics

Table 3 Comparison of different methods on the DREAM4 in-silico size 100 challenge networks

Method	iRafNet		HiDi		Winner		PFBNet
Data used	TS, KO		TS, KO		KO		TS, KO
Metrics	AUPR	AUROC	AUPR	AUROC	AUPR	AUROC	AUPR	AUROC
Network_1	0.552	0.901	0.63	0.916	0.536	0.914	0.639	0.935
Network_2	0.337	0.799	0.448	0.868	0.377	0.801	0.486	0.89
Network_3	0.414	0.835	0.413	0.797	0.39	0.833	0.469	0.886
Network_4	0.421	0.847	0.491	0.852	0.349	0.842	0.524	0.881
Network_5	0.298	0.792	0.251	0.803	0.213	0.759	0.448	0.866

The highest averaged AUPR and AUROC values are marked in bold for each network. TS, time-series expression data; KO, knockout data

ISSN: 1471-2105