Journal of Computer-Aided Molecular Design, 13: 167–183, 1999. KLUWER/ESCOM © 1999 Kluwer Academic Publishers. Printed in the Netherlands. 167 Comparison of two implementations of the incremental construction algorithm in flexible docking of thrombin inhibitors Ronald M.A. Knegtel a , Denis M. Bayada a , Richard A. Engh b , Wolfgang von der Saal c , Vincent J. van Geerestein a & Peter D.J. Grootenhuis a,* a Department of Molecular Design and Informatics, N.V. Organon, P.O. Box 20, 5340 BH Oss, The Netherlands; b Max-Planck-Institut für Biochemie, D-8152 Martinsried, Germany; c Boehringer Mannheim GmbH, Sandhofer Strasse 116, D-68305 Mannheim, Germany Received 15 May 1998; Accepted 28 August 1998 Key words: flexible docking, incremental construction algorithm, scoring function, structure-based drug design, thrombin Summary A set of 32 known thrombin inhibitors representing different chemical classes has been used to evaluate the per- formance of two implementations of incremental construction algorithms for flexible molecular docking: DOCK 4.0 and FlexX 1.5. Both docking tools are able to dock 10–35% of our test set within 2 Å of their known, bound conformations using default sampling and scoring parameters. Although flexible docking with DOCK or FlexX is not able to reconstruct all native complexes, it does offer a significant improvement over rigid body docking of sin- gle, rule-based conformations, which is still often used for docking of large databases. Docking of sets of multiple conformers of each inhibitor, obtained with a novel protocol for diverse conformer generation and selection, yielded results comparable to those obtained by flexible docking. Chemical scoring, which is an empirically modified force field scoring method implemented in DOCK 4.0, outperforms both interaction energy scoring by DOCK and the Böhm scoring function used by FlexX in rigid and flexible docking of thrombin inhibitors. Our results indicate that for reliable docking of flexible ligands the selection of anchor fragments, conformational sampling and currently available scoring methods still require improvement. Introduction Automated docking of rigid small molecules to their biomolecular receptors has proven to be a useful tool in the discovery of novel lead compounds of phar- maceutical relevance [1–4]. When large 3D databases of compounds are searched for possible lead com- pounds, only a limited amount of CPU time can be spent on each molecule. For this reason, most database searches by molecular docking have been performed with a single conformation for each ligand, thus reduc- ing the sampling problem to that of orienting a rigid molecule in a binding site. An interesting extension was recently reported by Lorber and Shoichet [5] in * To whom correspondence should be addressed. Present address: CombiChem Inc., 9050 Camino Santa Fe, San Diego, CA 92121, U.S.A. which ensembles of ligand conformers, superimposed on rigid fragments, were used for docking at compu- tational costs comparable to those required for single conformer docking. Several attempts have been made to include ligand flexibility in molecular docking, but until recently none of these approaches could per- form a thorough conformational search of the bound ligand in an amount of time suitable for database searching [6–9]. Substantially faster methods, based on the incremental construction algorithm of Leach and Kuntz [10], have recently been reported [11–13] and represent a more time-efficient approach to flex- ible docking. Currently, the only publicly available software intended for flexible docking of molecule databases are FlexX 1.5 [11] and DOCK 4.0 [14], which are two different implementations of the same incremental construction algorithm.