Table III.

Models with different quality water predictions for different category of docking models.

Listed are the numbers of models with bad, fair, good, and excellent predictions of water-mediated contacts (Columns 3-6) for each number of docking models ranked as incorrect, acceptable, medium or high quality in the classical CAPRI assessment (Columns 1,2).

	Docking Models	Models w/bad	Models w/fair	Models w/good	Models w/excellent
Incorrect	9	9	0	0	0
Acceptable	10	6	4	0	0
Medium	88	11	45	25	7
High	88	6	35	43	4