Predicting Mycobacterium tuberculosis Complex Clades Using Knowledge-Based Bayesian Networks

Table 2

Training 𝐹-measure for KBBN trained on all 10633 SITVIT isolates.

Clade	𝐹-measure	Clade	𝐹-measure	Clade	𝐹-measure
AFRI	0.800	H	0.736	PINI	0.750
AFRI_1	0.944	H1	0.924	PINI1	1.000
AFRI_2	0.908	H2	0.875	PINI2	0.667
AFRI_3	0.966	H3	0.915	S	0.976
Beijing	1.000	H3-Ural-1	0.873	T	0.926
BOV	0.948	H37Rv	0.958	T1-RUS2	0.778
BOV_1	0.993	H4-Ural-2	0.933	T2	0.953
BOV_2	1.000	LAM	0.947	T2-Uganda	0.991
BOV_3	0.644	LAM1	0.977	T3	0.964
BOV_4-Caprae	0.891	LAM11-ZWE	0.954	T3-ETH	0.65
Cameroon	0.929	LAM12-Madrid1	0.947	T3-OSA	0.626
CANETTI	1.000	LAM2	0.991	T4	0.988
CAS	0.937	LAM3	0.988	T4-CEU1	1.000
CAS1-Delhi	0.961	LAM4	0.970	T5	0.984
CAS1-Kili	0.973	LAM5	0.978	T5-Madrid2	1.000
CAS2	0.921	LAM6	0.856	T5-RUS1	0.949
EAI	0.982	LAM8	1.000	T-Tuscany	1.000
EAI1-SOM	0.986	Manu_ancestor	1.000	Turkey	0.928
EAI2-Manila	0.984	Manu1	0.991	X1	0.989
EAI2-Nonthaburi	1.000	Manu2	1.000	X2	0.963
EAI3-IND	0.963	Manu3	1.000	X3	0.995
EAI4-VNM	1.000	Microti	0.750	ZERO	0.800
EAI6-BGD1	0.989
EAI7-BGD2	1.000			AVERAGE	0.930
EAI8-MDG	1.000