Library

Notes: A new method for performing molecular dynamics simulations under constant pressure is presented. In the method, which is based on the extended system formalism introduced by Andersen, the deterministic equations of motion for the piston degree of freedom are replaced by a Langevin equation; a suitable choice of collision frequency then eliminates the unphysical "ringing'' of the volume associated with the piston mass. In this way it is similar to the "weak coupling algorithm'' developed by Berendsen and co-workers to perform molecular dynamics simulation without piston mass effects. It is shown, however, that the weak coupling algorithm induces artifacts into the simulation which can be quite severe for inhomogeneous systems such as aqueous biopolymers or liquid/liquid interfaces.

Notes: An improved understanding of the function of macromolecules of biological interest requires the characterization of their low-frequency internal fluctuations ((h-dash-bar)w(very-much-less-than)kBT where kBT∼200 cm−1 at room temperature), which dominate the atomic displacements. To explore the potential or inelastic neutron scattering for supplying information about such motions, we employ the formalism of Zemach and Glauber [A. C. Zemach and R. J. Glauber, Phys. Rev. 101, 118 (1956)] to calculate the one-phonon vibrational incoherent scattering function, Svibinc (q,w), of a small protein, BPTI. An in vacuo normal mode analysis of BPTI [B. Brooks and M. Karplus, Proc. Natl. Acad. Sci. U.S.A. 80, 6571 (1983)], is used for the calculation of Svibinc (q,w). The dependence of Svibinc (q,w) on energy transfer and momentum transfer is investigated and the results are placed in the context of present instrumental capabilities. Although the overall energy dependence of Svibinc (q,w) has a simple form, the variations in thehydrogen atom displacements in the different modes have a significant effect on the scattering. Mode anisotropy is strongly manifested in oriented spectra calculated without instrumental resolution broadening. Contributions from individual atoms and residues to the unbroadened whole molecule scattering are examined in detail. They are found to depend significantly on the atom or residue examined. This suggests that inelastic neutron scattering has considerable potential for the investigation of local dynamic variations in proteins by use of a deuterated protein with specifically protonated residues.Orientational averaging and instrumental resolution broadening of Svibinc (q,w) indicate that the one-phonon scattering detectable on the most suitable current instrument would show strong features arising principally from the frequency distribution of the vibrational modes. Multiphonon scattering is small at low momentum transfers in the low frequency range of most interest (〈30 cm−1) but increases rapidly with increasing energy and momentum transfers. It is shown to originate from the lowest frequency modes. The momentum transfer dependence of the one-phonon Svibinc (q,w) is investigated. An expression is found which can be used to extrapolate the calculated one-phonon scattering to zero momentum transfer using experimentally obtainable ranges of momentum transfer. This permits the extraction of an amplitude-weighted frequency distribution function. At current resolutions this function is similar in form to the unweighted frequency distribution function, knowledge of which would be very useful for analysis of protein thermodynamics and dynamics.

Notes: Statistical ensembles for simulating liquid interfaces at constant pressure and/or surface tension are examined, and equations of motion for molecular dynamics are obtained by various extensions of the Andersen extended system approach. Valid ensembles include: constant normal pressure and surface area; constant tangential pressure and length normal to the interface; constant volume and surface tension; and constant normal pressure and surface tension. Simulations at 293 K and 1 atm normal pressure show consistent results with each other and with a simulation carried out at constant volume and energy. Calculated surface tensions for octane/water (61.5 dyn/cm), octane/vacuum (20.4 dyn/cm) and water/vacuum (70.2 dyn/cm) are in very good agreement with experiment (51.6, 21.7, and 72.8 dyn/cm, respectively). The practical consequences of simulating with two other approaches commonly used for isotropic systems are demonstrated on octane/water: applying equal normal and tangential pressures leads to an instability; and applying a constant isotropic pressure of 1 atm leads to a large positive normal pressure. Both results are expected for a system of nonzero surface tension. Mass density and water polarization profiles in the liquid/liquid and liquid/vapor interfaces are also compared.

Notes: A uniform neutralizing background plasma is employed by default in standard Ewald calculations for net-charged systems. We show here that this plasma leads to serious artifacts in both system energy and pressure, which lead to unrealistic behavior. These artifacts are especially critical to simulations where either net charge or volume is allowed to change. To correct these problems we institute a net-charge correction term that consists of subtracting off the Ewald sum for a single particle with charge equal to the net charge of the full system and an optional Born or Poisson–Boltzmann term. This correction decreases pressure artifacts by three orders of magnitude and allows rapid energy convergence as the cell size increases. The correction term is general, in that it applies to systems of any shape or net charge. The net-charge correction was tested for systems in all three phases of matter: gas, solid, and liquid, and found to be markedly superior to standard Ewald in all three cases. In the gas phase, isolated molecular energies are quickly achieved as the cell size increases. In the solid phase the vaporization energy of the NaCl crystal is reproduced using free energy perturbation techniques where a single atom is removed. In the liquid phase, the solvation free energy of Na+ was investigated. It is demonstrated that both the net-charge correction and an energy term twice the size of a Born term is required to obtain optimal free energies of ionic hydration. © 1998 American Institute of Physics.

Notes: Abstract. Coupled advances in empirical force fields and classical molecular dynamics simulation methodologies, combined with the availability of faster computers, has lead to significant progress towards accurately representing the structure and dynamics of biomolecular systems, such as proteins, nucleic acids, and lipids in their native environments. Thanks to these advances, simulation results are moving beyond merely evaluating force fields, displaying expected structural fluctuations, or demonstrating low root-mean-squared deviations from experimental structures and now provide believable structural insight into a variety of processes such as the stabilization of A-DNA in mixed water and ethanol solution or reversible β-peptide folding in methanol. The purpose of this overview is to take stock of these recent advances in biomolecular simulation and point out some common deficiencies exposed in longer simulations. The most significant methodological advances relate to the development of fast methods to properly treat long-range electrostatic interactions, and in this regard the fast Ewald methods are becoming the de facto standard.