Biomolecular Force Field Parameterization via Atoms-in-Molecule Electron Density Partitioning

Area of Science:

• Computational chemistry
• Biomolecular modeling
• Drug design

Background:

• Molecular mechanics force fields rely on empirical parameters from small molecules, neglecting polarization in large systems.
• Current parametrization is labor-intensive and lacks accuracy for complex biomolecules.

Purpose of the Study:

• To develop a novel method for deriving force field parameters directly from quantum mechanical calculations.
• To incorporate polarization effects and reduce empirical parameterization for biomolecular modeling.

Main Methods:

• Utilized linear-scaling density functional theory (DFT) and atoms-in-molecule (AIM) electron density partitioning.
• Derived environment-specific charges and Lennard-Jones parameters from quantum mechanical calculations.

Main Results:

• Significantly reduced the number of empirical parameters required for force fields.
• Naturally incorporated polarization effects into charges and Lennard-Jones parameters.
• Demonstrated scalability to large systems, including proteins.

Conclusions:

• The new method offers a more accurate and efficient approach to biomolecular modeling and drug design.
• Validated by successful computation of hydration free energies, liquid properties, and binding free energies.