An efficient direct method for geometry optimization of large molecules in internal coordinates

A new efficient procedure has been developed for geometry optimization of large molecules using internal coordinates. The method stores only the nonzero elements of the large transformation matrices, in the spirit of direct methods in electronic structure theory. Matrix inversion has been replaced by iterative solution of linear systems of equations by the preconditioned conjugate gradient method. A new incomplete Cholesky preconditioner proved essential to accelerate the conjugate gradient procedure. The geometries of several alpha helical alanine polypeptides, up to 50 alanine units, have been optimized by the new method, using the SYBYL force field. For larger systems, the number of energy/gradient evaluations is reduced by a factor of 6–10, compared to Cartesian optimization, and the cost of the optimization is small compared to the energy calculation. We expect this method to be useful in molecular mechanics and in mixed quantum mechanics/molecular mechanics calculations.

[1]  M. Frisch,et al.  Using redundant internal coordinates to optimize equilibrium geometries and transition states , 1996, J. Comput. Chem..

[2]  R. Lord.  Vibrational spectra and structure : Volume 7, J.R. Durig, ed., Elsevier Scientific Publishing Co., Amsterdam, 1978, xv + 388 pages. Dfl. 146,00, $63.50. , 1976 .

[3]  Peter Pulay,et al.  Geometry optimization by direct inversion in the iterative subspace , 1984 .

[4]  Peter Pulay,et al.  Systematic AB Initio Gradient Calculation of Molecular Geometries, Force Constants, and Dipole Moment Derivatives , 1979 .

[5]  A. Komornicki,et al.  Rapid geometry optimization for semi-empirical molecular orbital methods , 1971 .

[6]  R. Abagyan,et al.  New methodology for computer-aided modelling of biomolecular structure and dynamics. 1. Non-cyclic structures. , 1989, Journal of biomolecular structure & dynamics.

[7]  Ruben Abagyan,et al.  ICM—A new method for protein modeling and design: Applications to docking and structure prediction from the distorted native conformation , 1994, J. Comput. Chem..

[8]  Peter Pulay,et al.  The calculation of ab initio molecular geometries: efficient optimization by natural internal coordinates and empirical correction by offset forces , 1992 .

[9]  W. Braun,et al.  Rapid calculation of first and second derivatives of conformational energy with respect to dihedral angles for proteins general recurrent equations , 1984, Comput. Chem..

[10]  Jon Baker,et al.  Geometry optimization in cartesian coordinates: The end of the Z‐matrix? , 1991 .

[11]  William H. Press,et al.  Numerical recipes , 1990 .

[12]  Jon Baker,et al.  The generation and use of delocalized internal coordinates in geometry optimization , 1996 .

[13]  Peter Pulay,et al.  Geometry optimization in redundant internal coordinates , 1992 .

[14]  P. Wormer,et al.  Conjugate gradient method for the solution of linear equations: Application to molecular electronic structure calculations , 1982 .

[15]  E. Bright Wilson,et al.  Some Mathematical Methods for the Study of Molecular Vibrations , 1941 .

[16]  Carl Eckart,et al.  Some Studies Concerning Rotating Axes and Polyatomic Molecules , 1935 .

[17]  Michael J. S. Dewar,et al.  Ground states of molecules. XXV. MINDO/3. Improved version of the MINDO semiempirical SCF-MO method , 1975 .

[18]  R. Cramer,et al.  Validation of the general purpose tripos 5.2 force field , 1989 .

[19]  Peter Pulay,et al.  Ab initio calculation of force constants and equilibrium geometries in polyatomic molecules , 1969 .

[20]  P. Pulay,et al.  AB Initio Vibrational Force Fields , 1984 .