7 - Exploring the conformational space: Distance geometry and model builders

Abstract

The three-dimensional geometry is an important property for understanding or predicting the behavior of molecular systems. It is the necessary starting point for the derivation of structural features, the estimation of steric requirements or the calculation of electronic properties. In many drug design applications, the “active analog” approach largely uses comparisons of molecular skeletons to deduce possible pharmacophores. Apart from direct experimental determinations (X-ray crystallography, for instance), molecular geometry can be attained either by database retrieval directly for structures already stored in the database or by assembling relevant substructures. Methods such as molecular mechanics are of prime interest to complement X-ray data and to constitute a privileged avenue to geometrical determinations not available from experimentation. The “distance geometry” approach works primarily on interatomic distances and allows the treatment of NOE enhancements in NMR. As stated by Crippen, “distance geometry refers to the study of geometric problems with an emphasis on distance among points.” The method utilizes a matrix of all-pairwise atomic distances in a molecule to generate a set of Cartesian coordinates consistent with this distance matrix. Distances come from experiments or are derived from standard geometrical features. The method is applicable to conformational search on small or medium size molecules but also appears very efficient to determine conformations of macromolecules in solution.