m5U-SVM: identification of RNA 5-methyluridine modification sites based on multi-view features of physicochemical features and distributed representation

Table 2 Performance comparison of different machine learning methods

Mode	Classifier	tenfold CV						Independent testing
Mode	Classifier	Acc (%)	Sn (%)	Sp (%)	MCC	F1	Acc (%)	Sn (%)	Sp (%)	MCC	F1
Full transcript	SVM	88.876	81.226	92.977	0.7527	0.8360	90.821	87.400	93.160	0.8091	0.8855
	RF	83.963	68.188	92.418	0.6384	0.7479	86.758	78.000	92.749	0.7239	0.8271
	LightGBM	86.669	76.466	92.138	0.7022	0.8001	89.439	84.000	93.159	0.7800	0.8659
	NB	79.731	81.029	79.035	0.5802	0.7362	80.909	82.400	79.891	0.6144	0.7781
	LR	85.873	76.662	90.810	0.6854	0.7911	86.758	80.600	90.971	0.7237	0.8317
	KNN	86.510	78.292	90.915	0.7003	0.8020	88.789	84.800	91.518	0.7667	0.860
	DT	76.410	65.710	82.145	0.4797	0.6565	80.016	76.000	82.763	0.5865	0.7554
	DL1	80.406	89.891	66.540	0.5906	0.7323	80.259	71.800	86.046	0.5869	0.7471
	DL2	83.883	90.684	73.940	0.6712	0.7827	84.971	82.000	87.004	0.6889	0.8159
Mature mRNA	SVM	94.358	92.981	95.736	0.8875	0.9402	94.106	93.061	95.142	0.8823	0.9402
	RF	92.429	90.234	94.619	0.8494	0.9163	91.869	89.387	94.331	0.8383	0.9163
	LightGBM	92.378	90.946	93.807	0.8479	0.9215	92.276	91.020	93.522	0.8457	0.9214
	NB	88.161	91.150	85.178	0.7646	0.8706	86.585	90.612	82.591	0.7342	0.8705
	LR	92.632	92.472	92.792	0.8527	0.9084	90.853	91.020	90.688	0.8171	0.9083
	KNN	91.514	94.608	88.426	0.8319	0.9040	90.243	92.244	88.259	0.8055	0.9040
	DT	86.280	86.673	85.888	0.7256	0.8775	87.804	87.755	87.854	0.7561	0.8775
	DL1	87.947	91.094	84.775	0.7629	0.8750	88.008	83.670	92.307	0.7638	0.8742
	DL2	89.858	85.632	94.049	0.8021	0.8934	91.260	86.530	95.951	0.8295	0.9079

ISSN: 1741-7007