Reproducing the conformations of protein-bound ligands: A critical evaluation of several popular conformational searching tools

Several programs (Catalyst, Confort, Flo99, MacroModel, and Omega) that are commonly used to generate conformational ensembles have been tested for their ability to reproduce bioactive conformations. The ligands from thirty-two different ligand–protein complexes determined by high-resolution (le2.0 Å) X-ray crystallography have been analyzed. The Low-Mode Conformational Search method (with AMBER* and the GB/SA hydration model), as implemented in MacroModel, was found to perform better than the other algorithms. The rule-based method Omega, which is orders of magnitude faster than the other methods, also gave reasonable results but were found to be dependent on the input structure. The methods supporting diverse sampling (Catalyst, Confort) performed least well. For the seven ligands in the set having eight or more rotatable bonds, none of the bioactive conformations were ever found, save for one exception (Flo99). These ligands do not bind in a local minimum conformation according to AMBER*\GB/SA. Taking these last two observations together, it is clear that geometrically similar structures should be collected in order to increase the probability of finding the bioactive conformation among the generated ensembles. Factors influencing bioactive conformational retrieval have been identified and are discussed.

[1]  Andrew Smellie,et al.  Poling: Promoting conformational variation , 1995, J. Comput. Chem..

[2]  Owen Johnson,et al.  The development of versions 3 and 4 of the Cambridge Structural Database System , 1991, J. Chem. Inf. Comput. Sci..

[3]  N. Metropolis,et al.  Equation of State Calculations by Fast Computing Machines , 1953, Resonance.

[4]  Z. Xu,et al.  The adipocyte lipid-binding protein at 1.6-A resolution. Crystal structures of the apoprotein and with bound saturated and unsaturated fatty acids. , 1994, The Journal of biological chemistry.

[5]  Peter A. Kollman,et al.  Many-body potential for molecular interactions , 1988 .

[6]  Gerhard Klebe,et al.  Comparison of Automatic Three-Dimensional Model Builders Using 639 X-ray Structures , 1994, J. Chem. Inf. Comput. Sci..

[7]  Colin McMartin,et al.  QXP: Powerful, rapid computer algorithms for structure-based drug design , 1997, J. Comput. Aided Mol. Des..

[8]  Edward P. Jaeger,et al.  Conformational searching methods for small molecules. III. Study of stochastic methods available in SYBYL and MACROMODEL , 1996, J. Comput. Chem..

[9]  S. Pickett,et al.  GRid-INdependent descriptors (GRIND): a novel class of alignment-independent three-dimensional molecular descriptors. , 2000, Journal of medicinal chemistry.

[10]  G. Chang,et al.  Macromodel—an integrated software system for modeling organic and bioorganic molecules using molecular mechanics , 1990 .

[11]  G J Williams,et al.  The Protein Data Bank: a computer-based archival file for macromolecular structures. , 1977, Journal of molecular biology.

[12]  Max Dobler,et al.  Multi-conformational Ligand Representation in 4D-QSAR: Reducing the Bias Associated with Ligand Alignment , 2000 .

[13]  Jonas Boström,et al.  Conformational energy penalties of protein-bound ligands , 1998, J. Comput. Aided Mol. Des..

[14]  Ralph E. Christoffersen,et al.  Computer-assisted drug design , 1979 .

[15]  Han van de Waterbeemd,et al.  Computer-Assisted Lead Finding and Optimization , 1997 .

[16]  W. C. Still,et al.  Semianalytical treatment of solvation for molecular mechanics and dynamics , 1990 .

[17]  Y. Martin,et al.  3D database searching in drug design. , 1992, Journal of medicinal chemistry.