How accurately do current force fields predict experimental peptide conformations? An adiabatic free energy dynamics...

Area of Science:

• Computational chemistry and biophysics
• Molecular dynamics simulations
• Spectroscopy analysis

Background:

• Classical biomolecular simulations face limitations in force field accuracy and sampling.
• Accurate conformational sampling is crucial for understanding peptide behavior.
• Existing methods for defining conformational states can be arbitrary.

Purpose of the Study:

• To address the challenge of comprehensive configuration space sampling in biomolecular simulations.
• To develop a robust method for defining and analyzing conformational macrostates of peptides.
• To assess the accuracy of modern biomolecular force fields in predicting peptide conformational preferences.

Main Methods:

• Employed driven adiabatic free energy dynamics for enhanced sampling.
• Utilized dihedral angles as collective variables to map free energy surfaces.
• Introduced a fuzzy clustering algorithm for objective macrostate delineation in collective-variable space.
• Calculated conformational populations for dipeptides using six different biomolecular force fields.

Main Results:

• Achieved converged free energy surfaces for dipeptides in both gas and solution phases.
• The fuzzy clustering algorithm successfully identified conformational macrostates without predefined boundaries.
• Calculated conformational preferences of dipeptides using six contemporary force fields.
• Compared calculated conformer populations against experimental Raman and IR-UV spectroscopy data.

Conclusions:

• The developed methodology effectively tackles the sampling problem in biomolecular simulations.
• Fuzzy clustering provides an objective approach to defining peptide conformational macrostates.
• The study offers a rigorous assessment of current biomolecular force field performance against experimental observations.