ParaMol: A Package for Automatic Parameterization of Molecular Mechanics Force Fields.

Area of Science:

• Computational Chemistry
• Molecular Modeling
• Drug Discovery

Background:

• Molecular mechanics (MM) simulations rely heavily on force field parameterization for accuracy.
• Parameterizing novel molecules, especially drug candidates with unique functional groups, remains a significant challenge.
• Existing force fields often lack accurate parameters for non-biomolecular compounds.

Purpose of the Study:

• To introduce ParaMol, a Python package designed for the parameterization of druglike molecules.
• To focus on fitting bonded and nonbonded terms using ab initio data.
• To streamline the parameterization process for molecular mechanics simulations.

Main Methods:

• Developed ParaMol, a Python package for automated force field parameterization.
• Utilized ab initio data fitting for bonded and nonbonded molecular mechanics terms.
• Tested methodologies on drug molecules like aspirin, caffeine, and a norfloxacin analogue.

Main Results:

• ParaMol successfully derived near-ideal bonded parameters for tested drug molecules.
• Demonstrated the software's ability to generate accurate parameters within functional form constraints.
• Analyzed the sensitivity of parameterization to different fitting datasets (dihedral scans, configurational ensembles) and weighting methods.

Conclusions:

• ParaMol provides a robust solution for parameterizing druglike molecules for MM simulations.
• The software can be integrated as a routine step in standard parameterization protocols.
• Accurate force field parameterization using ParaMol enhances the reliability of molecular simulations in drug discovery.