Andrey A. Bliznyuk

Simple method for locating possible ligand binding sites on protein surfaces

A new, fast, and easy‐to‐implement method, van der Waals–fast Fourier transform (vdW‐FFT), for locating possible binding sites on the surface of a protein was developed and tested on a set of 15 different enzyme–ligand complexes. The method scans the whole protein surface and possible ligand orientations in order to find the best geometrical match, which corresponds to the minimum of the modified vdW energy. Two different grids, fine and coarse, and two sets of MM parameters, from the OPLS and Amber‐94 force fields, were used. The method has been shown to work accurately on the fine grid. On the coarse grid, the vdW‐FFT method failed only on two complexes. The C program implementing the method and test set of proteins is available free on our web site: http://biocomp.anu.edu.au/~aab20: 983–988, 1999

Comparison of Linear-Scaling Semiempirical Methods and Combined Quantum Mechanical/Molecular Mechanical Methods for Enzymic Reactions. II. An Energy Decomposition Analysis.

QM/MM methods have been developed as a computationally feasible solution to QM simulation of chemical processes, such as enzyme‐catalyzed reactions, within a more approximate MM representation of the condensed‐phase environment. However, there has been no independent method for checking the quality of this representation, especially for highly nonisotropic protein environments such as those surrounding enzyme active sites. Hence, the validity of QM/MM methods is largely untested. Here we use the possibility of performing all‐QM calculations at the semiempirical PM3 level with a linear‐scaling method (MOZYME) to assess the performance of a QM/MM method (PM3/AMBER94 force field). Using two model pathways for the hydride‐ion transfer reaction of the enzyme dihydrofolate reductase studied previously (Titmuss et al., Chem Phys Lett 2000, 320, 169–176), we have analyzed the reaction energy contributions (QM, QM/MM, and MM) from the QM/MM results and compared them with analogous‐region components calculated via an energy partitioning scheme implemented into MOZYME. This analysis further divided the MOZYME components into Coulomb, resonance and exchange energy terms. For the model in which the MM coordinates are kept fixed during the reaction, we find that the MOZYME and QM/MM total energy profiles agree very well, but that there are significant differences in the energy components. Most significantly there is a large change (∼16 kcal/mol) in the MOZYME MM component due to polarization of the MM region surrounding the active site, and which arises mostly from MM atoms close to (<10 Å) the active‐site QM region, which is not modelled explicitly by our QM/MM method. However, for the model where the MM coordinates are allowed to vary during the reaction, we find large differences in the MOZYME and QM/MM total energy profiles, with a discrepancy of 52 kcal/mol between the relative reaction (product–reactant) energies. This is largely due to a difference in the MM energies of 58 kcal/mol, of which we can attribute ∼40 kcal/mol to geometry effects in the MM region and the remainder, as before, to MM region polarization. Contrary to the fixed‐geometry model, there is no correlation of the MM energy changes with distance from the QM region, nor are they contributed by only a few residues. Overall, the results suggest that merely extending the size of the QM region in the QM/MM calculation is not a universal solution to the MOZYME‐ and QM/MM‐method differences. They also suggest that attaching physical significance to MOZYME Coulomb, resonance and exchange components is problematic. Although we conclude that it would be possible to reparameterize the QM/MM force field to reproduce MOZYME energies, a better way to account for both the effects of the protein environment and known deficiencies in semiempirical methods would be to parameterize the force field based on data from DFT or ab initio QM linear‐scaling calculations. Such a force field could be used efficiently in MD simulations to calculate free energies.

Comparison of semiempirical and ab initio QM decomposition analyses for the interaction energy between molecules

Abstract A decomposition analysis of the interaction energy of molecular complexes using both semiempirical (PM3) and ab initio methods shows major differences. Whereas electrostatic stabilization accounted for a significant part of the interaction by ab initio theory, the electrostatic energy in semiempirical theory was mainly repulsive. This difference has major implications for intuitive models of intermolecular interactions, particularly in light of recent AM1 and PM3 energy decomposition calculations suggesting that charge transfer and polarization provides the binding energy of molecular clusters, including protein-solvent systems.

Identification and energetic ranking of possible docking sites for pterin on dihydrofolate reductase

The reliability of new methodology for detecting sites for ligand binding on the surfaces of proteins has been tested using a range of dihydrofolate reductase (DHFR) crystal structures. Docking of the pterin molecule to ten such DHFR structures has been examined. Initial docking sites were selected using the VDW-FFT method we have developed recently. This procedure was followed by rigid geometry optimization and solvation energy calculations using our parametrized reaction field multipoles (PRFM) method and the finite difference solution of the Poisson equation (FDPB) method. Two different sets of MM parameters, from the OPLS and Amber94 force fields, have been used. In eight cases the energy of the complexes with pterin bound at the active site was the lowest with the recent Amber94 parameters. In one case the spurious first-ranked site was only 1.8 kcal/mol lower in energy compared with the active site. The other ‘failure’ of the method may, in fact, represent a valid initial binding site. The calculations with the old OPLS parameters gave slightly worse results.

Comparison of electrostatic potential around proteins calculated from Amber and AM1 charges: application to mutants of prion protein

On the basis of arguments of complementary fit of shape and charge polarity or hydrophobicity, molecular electrostatic potentials (MEPs) around proteins are commonly used to deduce likely sites for interaction with ligands or other proteins, including for variations such as mutations. But protein MEPs calculated classically from fixed force field descriptions, including those with implicit solvent models such as in Delphi, do not allow for repolarization of protein residues within the protein system; hence, their representations are likely to be variably inaccurate. Linear-scaling methods now allow calculation of MEPs quantum mechanically for systems as large as proteins, and can account for polarization explicitly. Here we compare MEPs derived from AM1 charge distributions calculated by Mopac2000 with those from the classical Amber force field. Our models are mutants of prion protein (PrP), a protein with an unusually high number of charged residues. The results demonstrate that static point charges, as used in most current force fields, cannot reproduce the MEP of macromolecules. Also, it is not sufficient to account for the influence of nearby atoms connected by chemical bonds; the influence of nearby atoms in space is at least as important. Thus, further progress in the accuracy and wider applicability of force fields requires proper accounting for polarization. Mopac2000 calculations can provide the necessary data for checking new force fields and/or parameter fitting.

Computational chemistry on Fujitsu vector parallel processors: development and performance of applications software

In this and a preceding paper, we provide an introduction to the Fujitsu VPP range of vector-parallel supercomputers and to some of the computational chemistry software available for the VPP. Here, we consider the implementation and performance of seven popular chemistry application packages. The codes discussed range from classical molecular dynamics to semiempirical and ab initio quantum chemistry. All have evolved from sequential codes, and have typically been parallelised using a replicated data approach. As such they are well suited to the large-memory/fast-processor architecture of the VPP. For one code, CASTEP, a distributed-memory data-driven parallelisation scheme is presented

Faster gradients for semiempirical methods

A new algorithm for the numerical evaluation of gradients in semiempirical methods is described. The method is approximately twice as fast as the schemes currently employed and produces gradients of comparable accuracy. This method has been tested by comparing the results obtained by the new method with those of the previous numerical scheme, and also with those calculated analytically. The results of using the new gradients in geometry optimizations are also presented. ©1999 John Wiley & Sons, Inc. J Comput Chem 20: 629–635, 1999

Charge Transfer Effects in the GroEL–GroES Chaperonin Tetramer in Solution

In this work, we present the results of a large-scale, semiempirical LocalSCF quantum mechanical study of GroEL-GroES chaperonin in solution containing 2,481,723 atoms. We find that large biological systems exhibit strong quantum mechanical character, the extent of which was not previously known. Our data show that protein transfers -743 electron units of charge to solvent, which is not described by classical force fields. Contrary to the commonly held belief, which is based on classical mechanics, our computational data suggest that the quantum mechanical effects of charge transfer increase with the size of biological systems. We show that the neglect of charge transfer in classical force fields leads to significant error in the electrostatic potential of the macromolecule. These findings illustrate that a quantum mechanical framework is necessary for a realistic description of electrostatic interactions in large biological systems.