Table 2.

The performance of amino acid and dipeptide composition based method using different SVM kernels

Amine receptors	Amino acid composition based method				Dipeptide composition based method
	REF kernel (γ = 500 and C = 3)		Polynomial kernel (d = 1 and C = 2000)		RBF kernel (γ = 100 and C = 10)		Polynomial kernel (d = 1 and C = 1000)
	ACC	MCC	ACC	MCC	ACC	MCC	ACC	MCC
Acteylcholine	87.1	0.92	90.3	0.90	93.6	0.96	93.6	0.96
Adrenoceptor	95.5	0.88	86.3	0.76	100	0.93	100	0.91
Dopamine	92.1	0.82	84.2	0.74	92.1	0.95	92.0	0.93
Serotonin	85.3	0.85	83.3	0.85	98.2	0.97	94.4	0.95
Overall	89.8	0.86	85.6	0.81	96.4	0.95	95.1	0.93

ACC, accuracy.