TABLE 4.

Comparative results with traditional deep learning models.

Model	Sn	Sp	Pre	Acc	MCC	AUROC	AUPRC
CNN_model	24.1840	82.6800	84.6800	53.1360	0.1347	0.7768	0.8010
RNN_model	38.5520	81.5020	82.4500	59.8120	0.2688	0.7733	0.7966
ResNet_model	60.6560	60.7980	62.0140	60.7220	0.2150	0.6024	0.5957
PSATF-6mA (Decision Tree)	92.1600	93.2280	93.3620	92.6880	0.8543	0.9270	0.9474
PSATF-6mA (AdaBoost)	97.3980	94.7760	95.3220	96.1000	0.9232	0.9864	0.9797
PSATF-6mA (LightGBM)	99.4580	98.1960	98.2920	98.8360	0.9769	0.9988	0.9988
AttNet (XGBoost)	99.4940	98.7500	98.8060	99.1260	0.9826	0.9991	0.9991