Ligand-Based Virtual Screening Using Bayesian Inference Network and Reweighted Fragments

Table 5

The recall is calculated using the top 1% and top 5% of the DS2 data sets when ranked using the TAN, BIN, and BINRF.

Activity index	1%			5%
	TAN	BIN	BINRF	TAN	BIN	BINRF
07707	78.3**	72.18*	72.33*	91.08**	74.81*	74.17*
07708	74.01*	96*	100**	88.52*	99.61*	100**
31420	46.44*	79.82*	82.71**	77.6*	95.46*	97.15**
42710	57.22*	76.27*	95.36**	67.59*	92.55*	99.36**
64100	93.22*	88.43**	87.75*	97.89*	99.22**	98.93*
64200	63.39*	70.18*	71.79**	89.82*	99.2**	99.12*
64220	73.56*	68.32*	82.47**	92.05*	91.32*	98.89**
64500	60.75*	81.2*	96.56**	74.98*	94.96*	99.28**
64350	76.69*	81.89*	93.67**	90.34*	91.47*	98.24**
75755	95.99*	98.06*	98.26**	98.78**	98.33*	98.33*
Mean	71.957	81.235	88.09	86.86	93.69	96.34
Share cells	1	1	8	2	2	6