Improving pose prediction: the impact of flexibility Structure-based design, particularly docking, has had a profound impact on ligand design efforts in a great variety of projects. The greater part of this impact has come from the ability to predict reliably the likely pose of a new ligand in the binding site of the protein of interest (the so-called pose prediction problem). It is almost universally assumed, but rarely proved, that flexible approaches to pose prediction are superior to ones using only rigid models of the ligand and protein. This presentation will analyse the cost-benefit ratios of two approaches to docking available from OpenEye; one of them using a docking model entirely devoid of ligand or protein flexibility, the other incorporating increasing degrees of flexibility, both in the ligand and in the receptor model. The presentation will carefully assess the impact of increasing levels of flexibility, which will be balanced against any increase in computation time required. This analysis will allow a rational, informed selection of the best method for a particular problem in posing.