Free energy simulations: use of reverse cumulative averaging to determine the equilibrated region and the time required for convergence.

A method is proposed for improving the accuracy and efficiency of free energy simulations. The essential idea is that the convergence of the relevant measure (e.g., the free energy derivative in thermodynamic integration) is monitored in the reverse direction starting from the last frame of the trajectory, instead of the usual approach, which begins with the first frame and goes in the forward direction. This simple change in the use of the simulation data makes it straightforward to eliminate the contamination of the averages by contributions from the equilibrating region. A statistical criterion is introduced for distinguishing the equilibrated (production) region from the equilibrating region. The proposed method, called reverse cumulative averaging, is illustrated by its application to the well-studied case of the alchemical free energy simulation of ethane to methanol.

[1]  Arieh Ben-Naim,et al.  Solvation thermodynamics of nonionic solutes , 1984 .

[2]  T. Straatsma,et al.  Separation‐shifted scaling, a new scaling method for Lennard‐Jones interactions in thermodynamic integration , 1994 .

[3]  M Karplus,et al.  The meaning of component analysis: decomposition of the free energy in terms of specific interactions. , 1995, Journal of molecular biology.

[4]  G. Hummer.  Fast-growth Thermodynamic Integration∶ Results for Sodium Ion Hydration , 2002 .

[5]  M Karplus,et al.  Free energy simulations: The meaning of the individual contributions from a component analysis , 1994, Proteins.

[6]  Marcello Pagano,et al.  Principles of Biostatistics , 1992 .

[7]  C. Jarzynski.  Nonequilibrium Equality for Free Energy Differences , 1996, cond-mat/9610209.

[8]  T. Straatsma,et al.  Free energy of hydrophobic hydration: A molecular dynamics study of noble gases in water , 1986 .

[9]  Peter A. Kollman,et al.  FREE ENERGY CALCULATIONS : APPLICATIONS TO CHEMICAL AND BIOCHEMICAL PHENOMENA , 1993 .

[10]  J. Banavar,et al.  Computer Simulation of Liquids , 1988 .

[11]  H. Berendsen,et al.  Molecular dynamics with coupling to an external bath , 1984 .

[12]  C. Jarzynski,et al.  A “fast growth” method of computing free energy differences , 2001 .

[13]  M. Karplus,et al.  Simulation of activation free energies in molecular systems , 1996 .

[14]  Stefan Boresch,et al.  The Role of Bonded Terms in Free Energy Simulations. 2. Calculation of Their Influence on Free Energy Differences of Solvation , 1999 .

[15]  M Karplus,et al.  Specific amino acid recognition by aspartyl-tRNA synthetase studied by free energy simulations. , 1998, Journal of molecular biology.

[16]  Peter A. Kollman,et al.  A new method for carrying out free energy perturbation calculations: Dynamically modified windows , 1989 .

[17]  Peter A. Kollman,et al.  The lag between the Hamiltonian and the system configuration in free energy perturbation calculations , 1989 .

[18]  Olivier Michielin,et al.  Binding free energy differences in a TCR-peptide-MHC complex induced by a peptide mutation: a simulation analysis. , 2002, Journal of molecular biology.

[19]  Jan Hermans,et al.  Simple analysis of noise and hysteresis in (slow-growth) free energy simulations , 1991 .

[20]  Stefan Boresch,et al.  THE ROLE OF BONDED TERMS IN FREE ENERGY SIMULATIONS : 1. THEORETICAL ANALYSIS , 1999 .

[21]  Stefan Boresch,et al.  Influence of the treatment of electrostatic interactions on the results of free energy calculations of dipolar systems , 2000 .

[22]  M Karplus,et al.  The missing link between thermodynamics and structure in F1-ATPase , 2003, Proceedings of the National Academy of Sciences of the United States of America.

[23]  N. Doltsinis,et al.  Estimating pKa values for pentaoxyphosphoranes. , 2002, Journal of the American Chemical Society.

[24]  Thomas Simonson,et al.  A POISSON-BOLTZMANN STUDY OF CHARGE INSERTION IN AN ENZYME ACTIVE SITE : THE EFFECT OF DIELECTRIC RELAXATION , 1999 .

[25]  M. Karplus,et al.  CHARMM: A program for macromolecular energy, minimization, and dynamics calculations , 1983 .

[26]  J. Kirkwood.  Statistical Mechanics of Fluid Mixtures , 1935 .

[27]  D. Wallace,et al.  Statistical errors in molecular dynamics averages , 1985 .

[28]  R. Woods,et al.  Relative energies of binding for antibody-carbohydrate-antigen complexes computed from free-energy simulations. , 2000, Journal of the American Chemical Society.

[29]  S. Shapiro,et al.  An Analysis of Variance Test for Normality (Complete Samples) , 1965 .

[30]  R. Zwanzig.  High‐Temperature Equation of State by a Perturbation Method. I. Nonpolar Gases , 1954 .

[31]  David Chandler,et al.  Transition path sampling: throwing ropes over rough mountain passes, in the dark. , 2002, Annual review of physical chemistry.

[32]  W. L. Jorgensen,et al.  Comparison of simple potential functions for simulating liquid water , 1983 .

[33]  Jonathan W. Essex,et al.  Monte Carlo Simulations for Proteins: Binding Affinities for Trypsin−Benzamidine Complexes via Free-Energy Perturbations , 1997 .

[34]  Wei Yang,et al.  Generalized ensembles serve to improve the convergence of free energy simulations , 2003 .

[35]  R. Friedberg,et al.  Test of the Monte Carlo Method: Fast Simulation of a Small Ising Lattice , 1970 .

[36]  Bernard R. Brooks,et al.  New spherical‐cutoff methods for long‐range forces in macromolecular simulation , 1994, J. Comput. Chem..

[37]  P. Royston.  Approximating the Shapiro-Wilk W-test for non-normality , 1992 .

[38]  H. G. Petersen,et al.  Error estimates on averages of correlated data , 1989 .

[39]  Robert H. Wood,et al.  Estimation of errors in free energy calculations due to the lag between the hamiltonian and the system configuration , 1991 .

[40]  S. Shapiro,et al.  A Comparative Study of Various Tests for Normality , 1968 .

[41]  Thomas Simonson,et al.  Free energy simulations come of age: protein-ligand recognition. , 2002, Accounts of chemical research.