iPNHOT: a knowledge-based approach for identifying protein-nucleic acid interaction hot spots

BMC Bioinformatics

Table 3 Cross validation results of different classifiers based on the selected 7 features

Learning algorithms	REC	PRE	SPE	ACC	F1 score	MCC
KNN1	0.570	0.533	0.792	0.727	0.551	0.355
KNN3	0.512	0.595	0.855	0.754	0.550	0.384
KNN5	0.454	0.574	0.860	0.741	0.507	0.338
NB	0.384	0.579	0.884	0.737	0.462	0.308
LR	0.302	0.520	0.884	0.713	0.382	0.226
Random Forest^a	0.430	0.649	0.903	0.765	0.517	0.384
SVM (iPNHOT)	0.628	0.750	0.913	0.829	0.684	0.572

^aThe Random forest model is based on the all 97 features generated in this study, and the corresponding tree number is 68

ISSN: 1471-2105