A novel method to simulate the hydrophobic effect and its application to the ranking of host/guest complexes

Solvent entropy change is a major factor in driving the association of hydrophobic species in aqueous solutions. We have developed a novel methodology which simulates the solvation of hydrophobic surfaces by water. A system of virtual solvent particles surrounding the solute governed by arbitrarily determined rules provides a means to estimate the degree of order (Q) imposed by such solvation. Computed changes in Q (dQ) upon complex formation have been found to correlate well with observed binding affinities of host-guest complexes in aqueous solutions. Examples are described which illustrate the ability of dQ calculations to identify the correct ligand pose from a set of decoy complexes as well as provide rank ordering of a set of highly diverse ligand-protein complexes. Comparisons to surface-area-based calculations are discussed. The Q methodology holds great promise in the development of predictive structure-based approaches to drug design, as it provides a relatively simple means to estimate the hydrophobic effect.