ISIM: A Program for Grand Canonical Monte Carlo Simulations of the Ionic Environment of Biomolecules

Abstract

In this work we present a new software package (ISIM), which represents a flexible, computational tool for simulations of electrolyte solutions via a grand canonical Monte Carlo procedure (GCMC) with a specific capability of treating biomolecules in solution. The GCMC method provides a powerful tool for studying the ionic environments of highly charged macromolecules with attention to the atomic detail of both the solute and the mobile counterions. The ISIM software differs from previous schemes mainly by treating different ion types independently and offering a new parameterization procedure for calibrating excess chemical potentials and bulk ion concentrations. Additionally, ISIM leverages the APBS software package to provide accurate descriptions of the biomolecular electrostatic potential through the efficient solution of Poisson's equation. ISIM has been validated on a variety of test systems; we successfully reproduce elementary properties of electrolyte solutions as well as theoretical and experimental results for challenging test systems like Calmodulin and DNA.

Keywords:

• Grand canonical Monte Carlo
• Ion simulation
• ISIM software
• Electrostatics

Acknowledgements

The authors would like to thank P. Wolynes for helpful discussions on limitations of the Poisson-Boltzmann equation, W. Im and B. Roux for comments on grand canonical simulations, and H. Resat for advice on Monte Carlo methods. This work has been supported in part by grants from NIH, NSF, CTBP, NBCR, SDSC, and the W. M Keck Foundation.

Notes

*[email protected]

^‡[email protected]

^¶It is a problem that most HNC calculations use MC simulations as their own reference.

^§These references most often do not represent the latest version of the force-field though.

^∥This must not be confused with a re-interpretation in terms of parameterization though. It is impossible to get quantitative insight about possible corrections, as the concentration enters the algorithm in two distinct ways.

^#The structure we use is the one from Chattopadhyaya et al. the more recent one from Wilson et al. is less well-documented, as the publication has a slightly different focus.

^**This number was chosen, because the type of microscopic equilibrium dealt with here cannot be monitored via bulk quantities and might require a lot more random events.

^††There is a cleft in the middle an right pictures visible, which is due to missing interpolation in the graphics output and does not have any significance.