TOWARD A MORE EFFICIENT HANDLING OF CONFORMATIONAL FLEXIBILITY IN COMPUTER-ASSISTED MODELING OF DRUG MOLECULES

Several computational search techniques are described to map the confo rmation space of flexible organic molecules. A vast multiplicity of ge ometries is produced that has to be minimized according to a particula r energy function. Comparative studies on a nine-membered cyclic lacta m are taken as an example. They show that thoroughly tailored search c onditions can obtain roughly comparable search efficiencies. Out of th e vast multiplicity of geometrically possible and computationally acce ssible conformers, only a limited number will be of relevance for the problem under consideration. In ligand design for drug discovery, a re lative energy ranking determined on isolated conformers is only of lim ited use for the selection of biologically relevant conformers. This i s due to an unsatisfactory transferability of energy scales between di fferent energy functions and the strong modulation of conformational e nergies of isolated molecules once exposed to a structured molecular e nvironment. A knowledge-based approach, using torsion-angle libraries as retrieved for common fragments in small-molecule crystal structures , allows one to map more efficiently the biologically relevant part of conformation space. The relevance of these libraries for the conditio ns at the binding pocket of a protein is evidenced by experimental dat a. Sets of well-distributed conformers can be used to compare differen t drug molecules binding to common targets. Such comparisons reveal ne w modes of structural superposition of the molecules and consideration of their physicochemical properties leads to interesting pharmacophor e hypotheses. They indicate possible binding geometries at the recogni tion site of a protein and highlight the structural similarities and d ifferences that correlate with changes in the biological properties. C omparisons of GP IIb/IIIa receptor antagonists and of thrombin inhibit ors are discussed.