GPDOCK: highly accurate docking strategy for metalloproteins based on geometric probability

Area of Science:

• Biochemistry
• Computational Chemistry
• Structural Biology

Background:

• Predicting metalloprotein interactions is crucial for drug design and understanding enzymatic catalysis.
• The complexity of metal coordination in metalloproteins presents a significant challenge for existing docking methods.

Purpose of the Study:

• To develop a highly accurate docking method for metalloproteins.
• To improve the efficiency of structure-based drug design and metalloprotein mechanism analysis.

Main Methods:

• Developed Geometric Probability for Docking (GPDOCK), a novel docking method for metalloproteins.
• Utilized a structure-based machine learning model for scoring binding poses to enhance computational efficiency.
• Tested GPDOCK on 9360 metalloprotein-ligand complexes involving 10 common metal ions.

Main Results:

• GPDOCK achieved an accuracy of 94.3% in predicting metalloprotein-ligand binding poses.
• The method accurately handles docking scenarios involving water molecules in metal ion coordination.
• The integration of machine learning significantly improved computational efficiency.

Conclusions:

• GPDOCK offers an accurate and efficient solution for metalloprotein docking.
• This tool can significantly advance structure-based drug design and the study of metalloprotein binding mechanisms.
• The GPDOCK method and associated code are publicly available for further research.