CarSite-II: an integrated classification algorithm for identifying carbonylated sites based on K-means similarity-based undersampling and synthetic minority oversampling techniques

Table 2 Comparison of CarSite-II with other all available predictors on our independent test set

Classifier	TP	FP	Sn (%)	Sp (%)	Acc (%)	Mcc	AUC	G-mean
K
CarSPred.Y	36	547	30.77	92.65	91.69	0.1083	–	0.5339
CarSPred	5	89	4.27	98.80	97.34	0.0343	–	0.2055
CarSite	83	870	70.94	88.30	88.04	0.2203	0.8897	0.7915
CarSite-II	105	123	89.74	98.35	98.21	0.6358	0.9603	0.9395
P
CarSPred.Y	11	868	68.75	83.68	83.63	0.0773	–	0.7585
CarSPred	3	315	18.75	94.08	93.85	0.0296	–	0.4200
CarSite	9	587	56.25	88.96	88.86	0.0785	0.8671	0.7074
CarSite-II	13	108	81.25	97.97	97.92	0.2910	0.8768	0.8922
R
CarSPred.Y	14	387	25.93	93.51	92.91	0.0735	–	0.4924
CarSPred	10	321	18.52	94.62	93.94	0.0543	–	0.4186
CarSite	30	1236	55.56	79.28	79.07	0.0806	0.7715	0.6637
CarSite-II	43	112	79.63	98.12	97.96	0.4629	0.9236	0.8839
T
CarSPred.Y	8	335	33.33	94.85	94.63	0.0765	–	0.5623
CarSPred	4	271	16.67	95.84	95.54	0.0377	–	0.3997
CarSite	14	805	58.33	87.63	87.52	0.0840	0.7998	0.7150
CarSite-II	16	61	66.67	99.06	98.94	0.3685	0.8602	0.8127

ISSN: 1471-2105