Force Field for Peptides and Proteins based on the Classical Drude Oscillator

Area of Science:

• Computational chemistry
• Biophysics
• Molecular modeling

Background:

• Accurate molecular dynamics (MD) simulations are crucial for understanding peptide and protein behavior.
• Existing force fields may have limitations in capturing the nuances of molecular interactions and dynamics.
• Development of polarizable force fields aims to improve the physical realism of simulations.

Purpose of the Study:

• To present and validate a new polarizable force field, Drude-2013, for peptide and protein simulations.
• To optimize parameters for polypeptide backbones and side chain connectivity.
• To assess the model's performance in maintaining protein structure and capturing dynamic properties.

Main Methods:

• Development of a classical Drude oscillator-based polarizable force field.
• Parameterization using quantum mechanical calculations and experimental condensed phase data.
• Implementation in CHARMM and NAMD for molecular dynamics simulations.
• Validation through simulations of various peptides and proteins over 100 ns timescales.

Main Results:

• The Drude-2013 model maintains the folded state of peptides and proteins in explicit solvent simulations.
• It exhibits larger RMS differences and increased flexibility compared to the additive CHARMM36 force field.
• Improvements were observed in S² order parameters for specific residues, while NMR chemical shifts showed minor degradation.
• Significantly larger dipole moments were calculated for peptide backbones and tryptophan side chains.

Conclusions:

• The Drude-2013 polarizable force field provides a computationally accessible approach for more physically valid and internally consistent simulations of peptides and proteins.
• The model demonstrates potential for improved representation of molecular interactions and dynamics.
• Further refinements may address minor discrepancies observed in NMR data.