Table 2.

Comparison table of each evaluation metric for different models.

	AUC	AUPR	Pre	Rec	F1-Score
only-pre-training	0.9031	0.9071	0.8394	0.8582	0.8486
non-pre-training	0.9402	0.9409	0.8530	0.8967	0.8739
only-original feature	0.9422	0.9442	0.8920	0.9076	0.899
SGAEMDA	0.9585	0.9562	0.9037	0.9056	0.9046