Identification and characterization of plastid-type proteins from sequence-attributed features using machine learning

Table 8 Overall performance of various feature classifiers in 5-fold cross-validation for the classification of diverse plastid-types^* (phase-II)

Feature type	Sensitivity (%)	Specificity (%)	Accuracy (%)	MCC	Precision (%)	ER (%)	SVM kernel type
AAC	60.03	76.05	77.45	0.40	59.00	22.55	RBF (γ = 246, C = 1, j = 2)
PseAA	60.72	77.13	78.01	0.41	59.55	21.99	RBF (γ = 225, C = 1, j = 2)
Dipep	62.26	75.85	78.60	0.44	62.62	21.40	RBF (γ = 210, C = 1, j = 2)
NCC	60.97	77.34	78.39	0.42	58.51	21.61	RBF (γ = 5, C = 2, j = 3)
Phys-Chem	56.70	78.01	76.56	0.36	54.15	23.44	RBF (γ = 37, C = 9, j = 1)

*classification of 4 plastid types: chloroplast, chromoplast, etioplast, amyloplast; individual performance of these classifiers on each class can be seen in supplementary material; AAC = amino acid composition, PseAA = Pseudo amino acid composition, Dipep = Dipeptide composition, NCC = N_terminal-Center-C_terminal composition (sequence divided into 3 parts), Phys-Chem = Protein physicochemical properties, MCC = Matthews Correlation Coefficient, ER = Error Rate, RBF = Radial Basis Function of SVM.

ISSN: 1471-2105