Table 1.

The TPR, TNR and TDR of different methods

Scenario	JDINACa			DiffCorrb			DEDN			cPLR
	TDR	TPR	TNR	TDR	TPR	TNR	TDR	TPR	TNR	TDR	TPR	TNR
1	93.7	81.6	99.9	81.3	100	99.8	33.5	96.7	99	19.8	64.9	97.3
2	95.6	74.5	99.9	85	100	99.7	16.5	89.1	94.3	25.6	49.8	97.1
3	88.3	69.5	99.3	7.5	0.2	99.8	2.1	10.1	81.6	53.6	23.6	98.1
4	99.9	99.8	100	3.8	0.4	99.9	5.0	0.2	100	0.7	0.3	99.8

Average of 100 replications, %, the best performance is highlighted in bold.

^aPair $(G_i,G_j)$ was taken as differential edge in the network for JDINAC, when the differential dependency weight $w_{ij}\ge 4$.

^bSet to control the FDR = 0.1.