Table 1. The performance of CNN and CNN + LSTM based on 6mA-rice-Lv dataset under different CNN layers and kernel sizes.

	Model	CNN layers	Kernel size	SP (%)	SN (%)	ACC (%)	MCC	AUC
CNN	1-256-5	1	5	65.37	72.81	69.09	0.38	0.76
	1-256-8	1	8	84.00	65.56	74.78	0.51	0.84
	1-256-10	1	10	83.73	69.34	76.53	0.54	0.86
	1-256-16	1	16	79.95	86.85	83.40	0.67	0.91
	2-256-5	2	5	80.94	83.87	82.41	0.65	0.90
	2-256-8	2	8	70.33	94.13	82.23	0.67	0.92
	2-256-10	2	10	80.21	96.97	88.59	0.78	0.97
	2-256-16	2	16	91.98	94.53	93.26	0.87	0.98
	3-256-5	3	5	66.39	93.94	80.17	0.63	0.91
	3-256-8	3	8	82.14	96.89	89.52	0.80	0.97
	3-256-10	3	10	86.19	97.09	91.64	0.84	0.97
	3-256-16	3	16	86.88	96.85	91.86	0.84	0.98
CNN +LSTM	1-256-5-32	1	5	71.86	88.13	80.00	0.61	0.88
	1-256-8-32	1	8	89.82	95.02	92.42	0.85	0.97
	1-256-10-32	1	10	91.39	95.33	93.36	0.87	0.98
	1-256-16-32	1	16	91.91	94.93	93.42	0.87	0.98
	2-256-5-32	2	5	86.73	91.80	89.26	0.79	0.95
	2-256-8-32	2	8	92.63	94.67	93.65	0.87	0.98
	2-256-10-32	2	10	92.57	94.70	93.63	0.87	0.98
	2-256-16-32	2	16	92.19	95.11	93.65	0.87	0.98
	3-256-5-32	3	5	88.67	92.76	90.72	0.82	0.96
	3-256-8-32	3	8	92.49	93.56	93.03	0.86	0.97
	3-256-10-32	3	10	93.33	94.03	93.68	0.87	0.98
	3-256-16-32	3	16	93.09	94.55	93.82	0.88	0.98