Predicting blood-brain barrier permeation from three-dimensional molecularstructure

Predicting blood-brain barrier (BBB) permeation remains a challenge in drug design. Since it is impossible to determine experimentally the BBB partiti oning of large numbers of preclinical candidates, alternative evaluation me thods based on computerized models are desirable. The present study was con ducted to demonstrate the value of descriptors derived from 3D molecular fi elds in estimating the BBB permeation of a large set of compounds and to pr oduce a simple mathematical model suitable for external prediction. The met hod used (VolSurf) transforms 3D fields into descriptors and correlates the m to the experimental permeation by a discriminant partial least squares pr ocedure. The model obtained here correctly predicts more than 90% of the BB B permeation data. By quantifying the favorable and unfavorable contributio ns of physicochemical and structural properties, it also offers valuable in sights for drug design, pharmacological profiling, and screening. The compu tational procedure is fully automated and quite fast. The method thus appea rs as a valuable new tool in virtual screening where selection or prioritiz ation of candidates is required from large collections of compounds.