HawkRank: a new scoring function for protein-protein docking based on weighted energy terms

Area of Science:

• Computational biology
• Structural bioinformatics
• Biophysics

Background:

• Understanding protein-protein interactions (PPIs) is crucial for cell biology.
• Computational protein-protein docking aids experimental methods.
• Accurate scoring functions are vital for reliable docking predictions.

Purpose of the Study:

• Develop HawkRank, a novel scoring function for protein-protein docking.
• Improve the accuracy of identifying near-native protein complex structures.

Main Methods:

• Developed a solvation model based on solvent accessible surface area (SASA) and Generalized Born.
• Computed van der Waals and electrostatic potentials using the Amber ff14SB force field.
• Optimized weights for five energy terms to derive the HawkRank scoring function.

Main Results:

• HawkRank demonstrated superior performance compared to ZRANK, FireDock, and dDFIRE.
• The novel scoring function achieved higher scores based on the modified success rate (MSR).
• HawkRank effectively distinguishes near-native structures from decoys in protein-protein docking.

Conclusions:

• HawkRank is a highly effective scoring function for protein-protein docking.
• The developed scoring function enhances the accuracy of structural modeling for PPIs.
• HawkRank is publicly available for research use.