Evaluation of configurational entropy methods from peptide folding-unfolding simulation

Area of Science:

• Computational Biology
• Biophysics
• Protein Dynamics

Background:

• Understanding protein folding is crucial for molecular biology.
• Characterizing free energy surfaces and configurational entropy is key to studying protein dynamics.

Purpose of the Study:

• To characterize the free energy surface of a protein beta-hairpin.
• To evaluate the performance of different configurational entropy estimators.

Main Methods:

• 4-micros molecular dynamics simulation of Streptococcal protein G's B1 domain.
• Clustering of 200,000 conformers based on root-mean-square deviation (RMSD).
• Evaluation of relative free energies and configurational entropies using partition functions.

Main Results:

• Significant correlation found between free energy barrier height and pairwise RMSD.
• Quasi-harmonic entropy estimator in dihedral angle space outperformed Cartesian coordinate-based methods.
• A generalized quasi-harmonic approach showed comparable performance.
• High linear correlation (0.92-0.97) achieved with the best entropy estimators.

Conclusions:

• The study provides insights into the free energy landscape of protein folding.
• Dihedral angle-based entropy estimation is more effective for protein dynamics.
• Neglecting dihedral angle correlations unexpectedly improved agreement with reference entropies.