Transferable and Polarizable Coarse Grained Model for Proteins─ProMPT

Area of Science:

• Computational Chemistry
• Biophysics
• Molecular Modeling

Background:

• Classical molecular dynamics (MD) simulations at atomic resolution (fine-grained, FG) face computational limitations for large biomolecular systems.
• Existing coarse-grained (CG) force fields often lack transferability or fail to capture protein secondary/tertiary structure formation.
• Protein systems possess vast conformational spaces, challenging FG MD simulations' ability to explore dynamics.

Purpose of the Study:

• To develop a novel transferable coarse-grained (CG) force field for accurate protein simulations.
• To enable efficient exploration of large protein conformational spaces.
• To capture protein structural transitions with an explicit environmental representation.

Main Methods:

• Developed a CG force field using pseudoatoms representing chemical groups for modular biomolecular system construction.
• Incorporated electronic polarization responsive to environmental fluctuations, coupled to protein structural transitions.
• Parametrized nonbonded interactions using physics-based features (solvation/partitioning free energies) and bonded potentials from protein structure databases.

Main Results:

• Validated the CG model through simulations of diverse aqueous protein systems (Trp-cage, Trpzip4, villin, WW-domain, β-α-β).
• Demonstrated the force field's ability to preserve transferability across various environments and conditions.
• Successfully applied the force field to study the aqueous aggregation of Aβ 16-22 peptides.

Conclusions:

• The developed transferable CG force field offers an accurate and efficient alternative to FG MD for protein simulations.
• This CG model facilitates the study of complex biomolecular processes, including protein folding and aggregation.
• The explicit environment representation and polarization coupling enhance the simulation of protein dynamics and interactions.