CoDockPP: A Multistage Approach for Global and Site-Specific Protein-Protein Docking

Area of Science:

• Computational Biology
• Structural Biology
• Bioinformatics

Background:

• Protein-protein interactions are crucial for biological processes.
• Existing protein-protein docking methods face challenges in integrating global sampling with precise scoring.
• Efficient and accurate protein docking strategies are needed.

Purpose of the Study:

• To develop a novel multistage protein-protein docking strategy, CoDockPP.
• To enhance the accuracy and efficiency of protein-protein docking.
• To address limitations of existing Fast Fourier Transform (FFT)-based docking methods.

Main Methods:

• CoDockPP utilizes the sampling efficiency of FFT-based methods for initial pose selection.
• Retained poses undergo site constraint and atomic scoring in Cartesian space.
• Candidate poses are refined through local sampling and precise knowledge-based scoring.

Main Results:

• CoDockPP demonstrated significantly improved success rates and hit counts on the Protein-Protein Docking Benchmark 5.0.
• The method showed particular effectiveness in difficult *ab initio* and site-specific docking cases.
• CoDockPP outperforms previous strategies in predicting protein complex structures.

Conclusions:

• CoDockPP offers a powerful and accurate approach for protein-protein docking.
• The multistage strategy effectively combines global sampling with precise scoring.
• The CoDockPP server is publicly available for research use.