Abstract
The methodological dependence of observed ion-peptide associations in molecular dynamics simulations is investigated. We compare the results from several simulations of a pentapeptide in explicit solvent and salt ions which differ in the their treatment of the long ranged electrostatic interactions. Results for both the Ewald and switching function techniques are presented. It was found that there were important differences between the two methods for the water dipole-dipole temporal and spatial correlations, total dipole moment fluctuations, and self-diffusion constants. Electrostatic potentials calculated in the region of the peptide are also used to illustrate the large differences that can arise from different treatments of the electrostatic interactions. It appears that the switching function distorts the molecular electrostatic potential experienced by the salt ions to such a degree that their behaviour becomes highly dependent on the initial conditions. In summary, the use of a switching function is not recommended for the simulation of ions and their interactions with biomolecules.
Original language | English (US) |
---|---|
Pages (from-to) | 8430-8441 |
Number of pages | 12 |
Journal | Journal of Chemical Physics |
Volume | 95 |
Issue number | 11 |
DOIs | |
State | Published - 1991 |
Externally published | Yes |
ASJC Scopus subject areas
- General Physics and Astronomy
- Physical and Theoretical Chemistry