SOHSite: incorporating evolutionary information and physicochemical properties to identify protein S-sulfenylation sites

Table 2 Five-fold cross validation results for SVM models trained with various features individually

Training features	Sn	Sp	Acc	MCC
20D Binary code	0.66	0.68	0.68	0.23
BLOSUM62	0.68	0.70	0.69	0.26
Amino Acid Composition (AAC)	0.64	0.65	0.65	0.19
Amino Acid Pair Composition (AAPC)	0.64	0.67	0.67	0.21
Accessible Surface Area (ASA)	0.60	0.61	0.61	0.14
Secondary structure (SS)	0.56	0.56	0.56	0.08
Position Weight Matrix (PWM)	0.64	0.66	0.66	0.20
Position-specific scoring matrix (PSSM)	0.71	0.72	0.72	0.30

A total of 1145 positive data and 8368 negative data were used in the cross validation process. Sn, sensitivity; Sp, specificity; Acc, accuracy; MCC, Matthews Correlation Coefficient

ISSN: 1471-2164