Deepstacked-AVPs: predicting antiviral peptides using tri-segment evolutionary profile and word embedding based multi-perspective features with deep stacking model

Table 2 Predictive outcomes of the training dataset via different feature descriptors

Encoding method	Classifier	Acc (%)	Sn (%)	Sp (%)	MCC	AUC
CTDT	RF	88.07	90.90	85.71	0.76	0.95
	ETC	72.47	79.79	66.38	0.46	0.80
	XGB	82.11	78.78	84.87	0.63	0.90
	DNN	88.01	83.19	88.89	0.76	0.93
	Stacked-ensemble	89.90	93.93	86.55	0.80	0.95
PSSM-TS	RF	87.15	83.19	91.91	0.74	0.93
	ETC	87.61	84.87	90.91	0.75	0.94
	XGB	77.06	83.19	69.69	0.53	0.84
	DNN	86.23	89.07	82.82	0.72	0.91
	Stacked-ensemble	89.10	87.39	90.90	0.78	0.96
Word2vec (3mer)	RF	88.10	84.87	91.88	0.76	0.92
	ETC	74.31	78.99	68.61	0.47	0.76
	XGB	78.44	76.47	80.81	0.57	0.85
	DNN	87.61	84.48	91.90	0.76	0.92
	Stacked-ensemble	89.91	89.97	90.23	0.79	0.96
Hybrid vector	RF	88.91	85.71	92.92	0.78	0.94
	ETC	88.07	83.19	93.93	0.76	0.94
	XGB	87.61	88.23	86.85	0.75	0.93
	DNN	88.53	93.93	84.03	0.77	0.95
	Stacked-ensemble	92.20	90.75	93.91	0.84	0.96

ISSN: 1471-2105