Screened Electrostatic Interactions in Molecular Mechanics.

Area of Science:

• Computational Chemistry
• Molecular Modeling
• Physical Chemistry

Background:

• Molecular mechanics (MM) typically uses point charges for electrostatic interactions, which can be inaccurate at short distances.
• Existing MM models often compensate for electrostatic inaccuracies with empirical terms like Lennard-Jones potentials.

Purpose of the Study:

• To develop a more accurate electrostatic interaction model for molecular mechanics.
• To incorporate charge penetration effects into MM simulations using screened Coulomb potentials.

Main Methods:

• Developed a screened Coulomb potential scheme by reoptimizing screening parameters for MM.
• Utilized electrostatic-potential-fitted partial atomic charges for parameter optimization.
• Modeled atomic charge density as a sum of a point charge and a smeared charge distribution.

Main Results:

• The screened Coulomb potential model significantly improves electrostatic interaction accuracy compared to the point-charge model.
• Mean unsigned error in electrostatics decreased from 9.2 to 3.3 kcal/mol for a diverse dataset.
• Error in electrostatics for water dimers decreased from 1.7 to 0.5 kcal/mol.
• The screened MM method demonstrated higher accuracy than unscreened QM/MM calculations.

Conclusions:

• The new screened MM method provides a more physically realistic representation of electrostatic interactions.
• This approach offers a significant improvement over traditional point-charge models in molecular mechanics.
• The method is easily implementable in existing MM software, facilitating the development of improved force fields.