pyDockCG: new coarse-grained potential for protein-protein docking

Area of Science:

• Computational Biology
• Structural Biology
• Biophysics

Background:

• Protein-protein interactions are crucial for biological processes, necessitating accurate structural, functional, and energetic characterization.
• Computational docking methods predict protein complex structures, but struggle with conformational changes upon binding.
• Flexible docking is computationally expensive, limiting its widespread application.

Purpose of the Study:

• To develop an efficient coarse-grained (CG) potential, pyDockCG, for protein-protein docking scoring and refinement.
• To incorporate novel terms for Coulomb electrostatics and solvation energy into the CG model.
• To enable efficient treatment of flexibility in protein-protein docking.

Main Methods:

• Developed pyDockCG, a CG potential based on the UNRES model.
• Integrated new terms for Coulomb electrostatics and solvation energy, adapting the EEF1 model for CG.
• Applied pyDockCG to rigid-body docking sets for comparison with full-atom methods.

Main Results:

• pyDockCG achieved results comparable to the full-atom scoring function pyDock on rigid-body docking sets.
• The CG approach significantly reduced computational cost compared to full-atom methods.
• The new electrostatic and solvation terms improved scoring accuracy.

Conclusions:

• pyDockCG provides an efficient and accurate method for protein-protein docking, especially for complexes involving conformational changes.
• The CG approach offers a viable solution to the computational cost bottleneck in flexible docking.
• This tool has significant potential for biotechnological and therapeutic applications requiring structural characterization of protein complexes.