Accelerated Molecular Dynamics for Peptide Folding: Benchmarking Different Combinations of Force Fields and Explicit...

Area of Science:

• Computational Chemistry
• Biophysics
• Molecular Dynamics

Background:

• Predicting peptide secondary structure is crucial in molecular biology.
• Accelerated molecular dynamics (aMD) offers enhanced sampling for conformational analysis.
• Force fields and solvation models significantly impact simulation accuracy.

Purpose of the Study:

• To evaluate different accelerated molecular dynamics (aMD) protocols for predicting peptide secondary structure.
• To compare the performance of various force fields (ff99SB, ff14SB, ff19SB) and solvation models (TIP3P, OPC).
• To assess the influence of initial conformations (extended vs. misfolded) on prediction accuracy.

Main Methods:

• Utilized accelerated molecular dynamics (aMD) simulations.
• Tested combinations of three force fields (ff99SB, ff14SB, ff19SB) and two solvation models (TIP3P, OPC).
• Simulated eight peptides with varying secondary structures (helical, β-hairpin, disordered) from extended and misfolded starting points.

Main Results:

• All tested aMD protocols accurately predicted helical peptide structures.
• The ff19SB force field, with either TIP3P or OPC solvation, performed well for β-hairpins, showing a slight preference for TIP3P.
• The ff19SB/OPC combination demonstrated the best performance for intrinsically disordered peptides.
• The ff14SB/TIP3P combination exhibited a notable helical bias.

Conclusions:

• The ff19SB force field generally shows robust performance across different peptide types.
• The choice of solvation model (TIP3P vs. OPC) can influence secondary structure prediction accuracy, particularly for disordered peptides.
• Initial conformation plays a role in the prediction of β-hairpin structures.