SIMILARITY SEARCHING IN FILES OF 3-DIMENSIONAL CHEMICAL STRUCTURES - FLEXIBLE FIELD-BASED SEARCHING OF MOLECULAR ELECTROSTATIC POTENTIALS

Similarity measures based on the molecular electrostatic potential (ME P) have been used to quantify the degree of resemblance between pairs of rigid three-dimensional (3-D) molecules. This paper discusses the e ffect of including molecular flexibility on the similarities that are calculated using such measures in searches of large 3-D databases. Exp eriments with a graph-theoretic approach to MEP-based similarity searc hing demonstrate the limitations of using multiple conformations to re present the variations in an MEP that arise from torsional flexibility . Better results have been obtained with a genetic algorithm (GA) that has been developed for calculating the similarity between the MEPs of pairs of molecules, one or both of which is flexible. Experiments wit h two small files of flexible 3-D structures show that this GA identif ies significantly better alignments than does a comparable GA for sear ching rigid 3-D molecules.