ChemRxiv
These are preliminary reports that have not been peer-reviewed. They should not be regarded as conclusive, guide clinical practice/health-related behavior, or be reported in news media as established information. For more information, please see our FAQs.
1/1
2 files

A Simple Method for Including Polarization Effects in Solvation Free Energy Calculations When Using Fixed-Charge Force Fields: Alchemically Polarized Charges.

preprint
revised on 10.06.2020 and posted on 11.06.2020 by Braden Kelly, William Smith
The incorporation of polarizability in classical force-field molecular simulations is an ongoing area of research. We focus here on its application to hydration free energy simulations of organic molecules. In contrast to computationally complex approaches involving the development of explicitly polarizable force fields, we present herein a simple methodology for incorporating polarization into such simulations using standard fixed-charge force-fields, which we call the Alchemically Polarized Charges (APolQ) method. APolQ employs a standard classical alchemical free energy change simulation to calculate the free energy difference between a fully polarized solute particle in a condensed phase and its unpolarized state in a vacuum. One electronic structure (ES) calculation to of the electron densities is required for each state: for the former, we use a Polarizable Continuum Model (PCM), and for the latter we use vacuum-phase electronic structure calculations.

We applied APolQ to hydration free energy data for a test set of 45 neutral solute molecules in the FreeSolv database, and compared results obtained using three different water models (SPC/E, TIP3P, OPC3) and using MBIS and RESP partial charge methodologies. ES calculations were carried out at the MP2 level of theory and with cc-pVTZ and aug-cc-pVTZ basis sets. In comparison with AM1-BCC, we found that APolQ outperforms it for the test set. Despite our method using default GAFF parameters, the MBIS partial charges yield Absolute Average Deviations (AAD) 1.5 to 1.9 kJ·mol−1 lower than AM1-BCC.

We conjecture that this method can be further improved by fitting the Lennard-Jones and torsional parameters to partial charges derived using MBIS or RESP methodologies.

Funding

Natural Sciences and Engineering Research Council of Canada (Strategic Program grant no. STPGP 479466-15)

History

Email Address of Submitting Author

bkelly08@uoguelph.ca

Institution

University of Guelph

Country

Canada

ORCID For Submitting Author

0000-0003-3708-6517

Declaration of Conflict of Interest

no conflict of interest

Version Notes

Version 2.

Exports