Refinement of Docked Protein-Protein Complexes Using Repulsive Scaling Replica Exchange Simulations

Area of Science:

• Computational biology
• Structural biology
• Biophysics

Background:

• Accurate protein-protein complex structure prediction is crucial for understanding cellular functions.
• Current methods like deep learning and docking often require further refinement and rescoring.
• Standard molecular dynamics (MD) simulations are computationally expensive and may not effectively improve docking results.

Purpose of the Study:

• To develop and present a novel simulation scheme for refining and evaluating protein-protein complex structures.
• To introduce a method that simultaneously improves structural accuracy and provides realistic free energy scoring.
• To detail the setup and application of the new simulation technique.

Main Methods:

• Development of a replica-exchange-based scheme with repulsive biasing (RS-REMD).
• Application of bias specifically to intermolecular interactions by increasing van der Waals radii.
• Utilizing multiple replicas with varying levels of repulsive bias between interacting proteins.

Main Results:

• RS-REMD effectively refines predicted protein-protein complex structures.
• The method allows for simultaneous and realistic free energy scoring of complexes.
• Demonstrated application on two biological examples.

Conclusions:

• RS-REMD offers a significant improvement over standard MD for protein complex refinement and scoring.
• This method provides a more efficient and accurate approach to evaluating predicted protein-protein interactions.
• The described technique and associated scripts facilitate further research in structural interactomics.