iIL13Pred: improved prediction of IL-13 inducing peptides using popular machine learning classifiers

BMC Bioinformatics

Table 2 Performance metrics of seven machine learning models for prediction of IL-13 inducing peptides on the top 10 features via mRMR (iIL13Pred) and its comparison with the state of the art (IL13Pred)

Classifier	Dataset	Sensitivity		Specificity		Accuracy		AUCROC		MCC
Classifier	Dataset	Proposed study	Jain et al. [19]	Proposed study	Jain et al. [19]	Proposed study	Jain et al. [19]	Proposed study	Jain et al. [19]	Proposed study	Jain et al. [19]
DT	Training	64.80	69.36	72.18	69.46	71.47	69.45	0.72	0.74	0.24	0.24
DT	Validation	65.08	60.26	76.29	71.43	75.19	70.08	0.75	0.72	0.27	0.22
GNB	Training	72.00	71.06	70.98	68.18	71.08	68.44	0.79	0.74	0.29	0.24
GNB	Validation	68.25	64.10	72.34	66.31	71.94	66.05	0.75	0.73	0.26	0.21
KNN	Training	68.40	65.11	66.17	56.47	66.38	57.26	0.72	0.64	0.21	0.13
KNN	Validation	65.08	60.26	69.93	55.73	69.46	56.28	0.70	0.61	0.22	0.11
LR	Training	66.00	64.26	62.59	61.85	62.92	62.07	0.67	0.67	0.17	0.15
LR	Validation	60.32	56.41	62.03	59.61	61.86	59.23	0.63	0.63	0.14	0.11
SVC	Training	60.40	54.47	74.46	70.70	73.10	69.22	0.74	0.67	0.23	0.16
SVC	Validation	49.21	46.15	79.04	74.25	76.12	70.85	0.71	0.64	0.20	0.15
RF	Training	76.40	74.47	75.02	76.34	75.16	76.17	0.84	0.83	0.34	0.33
RF	Validation	68.25	64.10	78.18	75.13	77.21	73.80	0.80	0.77	0.31	0.28
XGB	Training	79.60	78.30	73.99	74.84	74.53	75.16	0.84	0.83	0.34	0.33
XGB	Validation	74.60	71.80	75.77	73.02	75.66	72.87	0.83	0.80	0.33	0.30

ISSN: 1471-2105