Predicting lncRNA–protein interactions through deep learning framework employing multiple features and random forest algorithm

Table 2 Comparison of prediction results with different feature combinations on RPI1807

Combination of Features	ACC (%)	SEN (%)	SPE (%)	PPV (%)	F1 (%)	MCC	AUC (%)
PPI,PC	94.17	97.09	91.25	92.94	89.15	0.9465	99.31
Str, PPI	96.40	94.43	98.37	98.34	92.90	0.9633	99.14
Str, PC	97.47	95.72	99.23	99.21	95.01	0.9743	99.52
Seq, PPI	97.39	95.89	98.88	98.85	94.81	0.9735	99.17
Seq, PC	96.87	96.06	97.68	97.66	93.76	0.9685	99.25
Seq, Str	96.87	95.20	98.54	98.50	93.81	0.9682	99.22
Str, PPI, PC	97.56	95.71	99.40	99.39	95.19	0.9751	99.41
Seq, PPI, PC	94.08	96.23	91.93	93.59	88.93	0.9456	98.85
Seq, Str, PPI	97.60	96.10	99.31	99.29	95.51	0.9769	99.52
Seq, Str, PC	97.51	95.89	99.14	99.12	95.08	0.9747	99.48
Seq, Str, PPI, PC	97.73	95.72	99.49	99.47	95.27	0.9755	99.57

Bold values represent the maximum value of the corresponding evaluation indicator

ISSN: 1471-2105