Table 2.

The comparison among KinasePhos 2.0, DISPHOS, PredPhospho, GPS, PPSP and KinasePhos 1.0

Tools	DISPHOS	PredPhospho	GPS	PPSP	KinasePhos 1.0	KinasePhos 2.0
Method	Logistic regression	SVM	MCL+GPS	BDT	MDD+HMM	CP+SVM
Number of kinases	–	4 groups	71 groups	68 groups	18	58
Kinase PKA	–	Sn = 0.88	Sn = 0.89	Sn = 0.90	Sn = 0.91	Sn = 0.92
		Sp = 0.91	Sp = 0.91	Sp = 0.92	Sp = 0.86	Sp = 0.89
Kinase PKC	–	Sn = 0.79	Sn = 0.82	Sn = 0.82	Sn = 0.80	Sn = 0.84
		Sp = 0.86	Sp = 0.83	Sp = 0.86	Sp = 0.87	Sp = 0.86
Kinase CK2	–	Sn = 0.84	Sn = 0.83	Sn = 0.83	Sn = 0.87	Sn = 0.87
		Sp = 0.96	Sp = 0.88	Sp = 0.90	Sp = 0.85	Sp = 0.86
Serine	Acc = 0.76	Acc = 0.81	–	–	Acc = 0.86	Acc = 0.90
Threonine	Acc = 0.81	Acc = 0.77	–	–	Acc = 0.91	Acc = 0.93
Tyrosine	Acc = 0.83	–	–	–	Acc = 0.84	Acc = 0.88
Histidine	–	–	–	–	–	Acc = 0.93
Overall performance	–	Acc = 0.76 ∼ 0.91	–	–	Acc = 0.87	Acc = 0.91

SVM, support vector machine; MCL, Markov cluster algorithm; GPS, group-based phosphorylation scoring method; BDT, Bayesian decision theory; MDD, maximal dependence decomposition; HMM, hidden Markov model; CP, coupling pattern; Sn, sensitivity; Sp, specificity; Acc, accuracy.