Molecular dynamics simulations of protein folding

Area of Science:

• Computational Biology
• Biophysics
• Protein Folding

Background:

• Protein folding is crucial for biological function, but simulating it accurately is computationally challenging.
• Enhanced sampling methods are needed to overcome the long timescales associated with protein folding simulations.
• Replica exchange molecular dynamics (REMD) is a powerful technique for improving configurational sampling.

Purpose of the Study:

• To demonstrate the utility of the REMD algorithm for studying protein folding.
• To assess the accuracy of current force fields and enhanced sampling methods in predicting protein structures.
• To simulate the folding of a small protein, protein A, a 57-amino acid three-helix bundle.

Main Methods:

• Employed the replica exchange molecular dynamics (REMD) algorithm with multiple replicas at different temperatures.
• Utilized the Amber force field for molecular dynamics simulations.
• Performed simulations in explicit solvent, starting from extended protein configurations.

Main Results:

• Achieved stable configurations of protein A with backbone root mean square distance (RMSD) within 0.17 nm of the NMR-determined structure.
• Demonstrated that REMD significantly enhances configurational sampling, enabling faster folding studies.
• Obtained results without biasing the simulation towards the known folded structure.

Conclusions:

• REMD simulations, combined with the Amber force field, can accurately predict the native structure of small proteins like protein A.
• The study validates the performance of current computational tools for protein structure prediction.
• Enhanced sampling methods like REMD are effective for studying protein folding dynamics within accessible simulation times.