Parallel cascade selection molecular dynamics to screen for protein complexes generated by rigid docking.

Area of Science:

• Computational Biology
• Biophysics
• Structural Biology

Background:

• Rigid docking simulations provide initial protein complex models but often lack conformational flexibility.
• Enhanced sampling methods are needed to explore conformational landscapes and refine docking poses.
• Identifying biologically relevant protein-protein interactions requires accurate prediction of complex structures.

Purpose of the Study:

• To develop and validate a flexible docking simulation approach by combining rigid docking with parallel cascade selection molecular dynamics (PaCS-MD).
• To utilize PaCS-MD for reproducing protein association processes and screening generated decoys.
• To demonstrate the method's capability in identifying near-native protein complexes.

Main Methods:

• A hybrid approach integrating rigid docking with PaCS-MD for flexible docking simulations.
• PaCS-MD cycle involving selection of important initial structures and conformational resampling.
• Definition and monitoring of root-mean-square deviation from decoy (RMSD_decoy) to screen complexes.
• Application to the dimerization of K48-linked ubiquitin dimer.

Main Results:

• PaCS-MD successfully reproduced association processes from separated proteins to decoy structures.
• RMSD_decoy monitoring enabled the screening of non-near-native protein complexes.
• The method effectively filtered out irrelevant decoys generated by initial rigid docking.
• Near-native complexes were identified by imposing an RMSD_decoy threshold.

Conclusions:

• The proposed flexible docking simulation using PaCS-MD is an effective post-processing step for rigid docking.
• PaCS-MD can accurately identify and screen near-native protein complexes based on conformational sampling.
• This method enhances the reliability of protein-protein interaction structure prediction.