Table 3.

Performance comparison between our method and existing methods with different essential to non-essential gene ratios on InWeb_IM dataset.

Methods	Precision	Recall	SP	NPV	F-measure	MCC	ACC	AUC	AP
DeepWalk (1:6)	0.749	0.515	0.971	0.923	0.610	0.571	0.906	0.904	0.667
LINE (1:6)	0.819	0.165	0.994	0.877	0.275	0.332	0.875	0.832	0.547
Centrality (1:6)	0.527	0.562	0.916	0.926	0.544	0.465	0.865	0.828	0.511
Z-curve (1:6)	0.485	0.462	0.918	0.911	0.473	0.388	0.853	0.840	0.456
Our method (1:6)	0.841	0.550	0.983	0.929	0.665	0.641	0.921	0.915	0.762
DeepWalk (1:1)	0.817	0.848	0.811	0.842	0.832	0.659	0.829	0.908	0.894
LINE (1:1)	0.641	0.905	0.492	0.839	0.750	0.436	0.699	0.804	0.792
Centrality (1:1)	0.816	0.713	0.839	0.745	0.761	0.557	0.776	0.851	0.841
Z-curve (1:1)	0.730	0.777	0.713	0.762	0.753	0.491	0.745	0.827	0.801
Our method (1:1)	0.855	0.858	0.854	0.858	0.857	0.713	0.856	0.928	0.921