VoteDock: consensus docking method for prediction of protein-ligand interactions

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Molecular recognition is crucial for biological processes, making protein-ligand interaction prediction vital for drug discovery.
• Current in silico methods for screening molecular libraries and predicting protein-ligand poses are often time-consuming and expensive.

Purpose of the Study:

• To develop and present a novel consensus approach for accurately predicting both protein-ligand complex structures and binding affinities.
• To improve the efficiency and accuracy of identifying potential drug candidates.

Main Methods:

• A consensus approach integrating results from seven widely used docking programs: Surflex, LigandFit, Glide, GOLD, FlexX, eHiTS, and AutoDock.
• Evaluation on an extensive benchmark dataset of 1300 protein-ligand pairs from the PDBbind database.

Main Results:

• The consensus method improved proper docking by approximately 20% on average compared to individual docking programs.
• Achieved over 10% improvement in docking accuracy compared to the best single program.
• Reduced the Root Mean Square Deviation (RMSD) of predicted complex conformations by 0.5 Å.
• Increased the Pearson correlation of predicted binding affinity with experimental values to 0.5.

Conclusions:

• The novel consensus approach offers a significant advancement in predicting protein-ligand complex structures and binding affinities.
• This method enhances the accuracy and efficiency of virtual screening in drug discovery pipelines.
• The improved prediction accuracy can accelerate the identification of novel therapeutic agents.