Extended molecular dynamics and optimized Rouse–Zimm model studies of a short peptide: Various friction approximations

Abstract

Developing a theory for the long time dynamics of polypeptides requires not only a proper choice of the relevant dynamic variables, but also a meaningful definition of friction coefficients for the individual atoms or groups of atoms in the reduced system. We test various aspects of the optimized Rouse–Zimm model for describing the long time rotational dynamics of a peptide fragment. The necessary equilibrium input information is constructed from a 1 ns molecular dynamics simulation for the solvated peptide by using a parallel Cray version of CHARMm, whose new features are described here. The simulations also provide time autocorrelation functions for comparisons with both theoretical predictions and with experiment. Two atomic friction models (van der Waals radii and accessible surface area) are chosen, and tests are made of the applicability of two combining rules for calculating the group friction coefficients. While the rotational dynamics of the peptide is insensitive to the friction models used, the combining rules are found to impact profoundly upon the theoretical descriptions for the behavior of the peptide dynamics for the reduced descriptions with fewer variables. The calculations study the role of the memory functions, neglected in this dynamical theory, and the interatomic hydrodynamic interactions in constructing the group friction coefficients. While the 1 ns trajectory is longer than customarily used for very complex systems, there are nontrivial influences of the duration of the molecular dynamics trajectory on the description of the dynamics.

Long-Time Dynamics of Met-Enkephalin: Comparison of Theory with Brownian Dynamics Simulations

“…More recently, the GRZ approach has been applied to peptides containing a single tryptophane residue, glucagon, the hormone ACTH, and their fragments (Chen et al, 1987;Hu et al, 1990Hu et al, , 1991Hu et al, , 1995. The equilibrium averages required as inputs to the dynamical theory are obtained from realistic potentials, either with rotational isomeric state calculations, using the ECEPP potentials developed by Scheraga and co-workers (Li and Scheraga, 1987) or from MD simulations with the CHARMm package.…”

Mode-coupling Smoluchowski dynamics of a double-stranded DNA oligomer

“…This model of the additivity of group friction coefficients ignores the influence of hydrodynamic interactions within the group as discussed else- where, 24,25 thus overestimating the total friction in the macromolecule.…”

Maximum-Correlation Mode-Coupling Approach to the Smoluchowski Dynamics of Polymers

“…It has been shown that in the framework of ORZ dynamics all the time correlation functions describing the “local dynamics” can be derived exactly. , This reduced approach to local dynamics (ORZLD) uses a hydrodynamic description of the solvent, selects only part of the total degrees of freedom of the polymer (bonds or effective bonds) as “slow variables”, requires knowledge of the equilibrium averages of the bond vector correlation matrices and of the inverse bead distances, and models the friction coefficients of the atoms or group of atoms, “beads”, by the van der Waals additive method or the accessible surface area (ASA) method . The most advanced models in the ORZLD hierarchy 10 calculate the input averages by a time average along an MD trajectory, taking into account non-nearest-neighbor interactions described by empirical force fields, together with the solvent effect. , The effects on local dynamics of the length of the trajectory 13 and of the choice of bead friction coefficients were discussed. , These tests on the ORZ extended models were performed to observe the effect of progressively increasing the degrees of freedom included into the slow variables to reduce the effect of the discarded memory function, improving the description of the friction coefficients of “atoms” and of the “group of atoms”, and increasing the accuracy in the description of the interactions between the solute “atoms” and between solute and solvent.…”

“…12 The most advanced models in the ORZLD hierarchy 10 calculate the input averages by a time average along an MD trajectory, taking into account non-nearest-neighbor interactions described by empirical force fields, together with the solvent effect. 12,13 The effects on local dynamics of the length of the trajectory 13 and of the choice of bead friction coefficients were discussed. 12,14 These tests on the ORZ extended models were performed to observe the effect of progressively increasing the degrees of freedom included into the slow variables to reduce the effect of the discarded memory function, improving the description of the friction coefficients of "atoms" and of the "group of atoms", and increasing the accuracy in the description of the interactions between the solute "atoms" and between solute and solvent.…”