Table 5.

The prediction results compared with those of other methods on the training dataset.

Layer	Method	$\mathbf{Acc}$	$\mathbf{Sn}$	$\mathbf{Sp}$	$\mathbf{MCC}$
First Layer (Enhancer Identification)	iEnhancer-2L [19]	0.769	0.781	0.759	0.540
	iEnhancer-PsedeKNC [20]	0.768	0.773	0.763	0.540
	EnhancerPred [21]	0.732	0.726	0.738	0.464
	iEnhancer-EL [22]	0.780	0.757	0.804	0.561
	iEnhancer-5Step [23]	0.823	0.811	0.835	0.650
	iEnhancer-ECNN [24]	0.769	0.785	0.752	0.537
	iEnhancer-CNN [25]	0.806	0.759	0.889	0.693
	iEnhancer-XG [26]	0.811	0.757	0.865	0.627
	iEnhancer-GAN [This Study]	0.951	0.951	0.951	0.902
Second Layer (Enhancer Strength Identification)	iEnhancer-2L [19]	0.619	0.622	0.618	0.240
	iEnhancer-PsedeKNC [20]	0.634	0.626	0.644	0.270
	EnhancerPred [21]	0.621	0.627	0.615	0.241
	iEnhancer-EL [22]	0.650	0.690	0.611	0.315
	iEnhancer-5Step [23]	0.681	0.753	0.608	0.370
	iEnhancer-ECNN [24]	0.678	0.791	0.564	0.368
	iEnhancer-CNN [25]	0.764	0.436	0.768	0.451
	iEnhancer-XG [26]	0.667	0.749	0.586	0.340
	iEnhancer-GAN [This Study]	0.872	0.873	0.871	0.744