Properties that rank protein:protein docking poses with high accuracy

Area of Science:

• Computational Biology
• Structural Biology
• Biophysics

Background:

• Accurate prediction of protein:protein complex structures is crucial for molecular biology and drug discovery.
• Existing docking algorithms require robust methods for ranking predicted poses to identify near-native structures.

Purpose of the Study:

• To evaluate the effectiveness of interfacial area and MM-PBSA derived properties in ranking protein:protein docking poses.
• To assess the success rate of these methods in predicting high-quality near-native structures.

Main Methods:

• Utilized a dataset of 48 protein:protein complexes across various docking scenarios (bound:bound, bound:unbound, unbound:unbound).
• Applied Molecular Mechanics-Poisson Boltzmann Surface Area (MM-PBSA) calculations to estimate binding free energy of protein monomers.
• Scored docking poses using MM-PBSA and computational alanine scanning mutagenesis data.

Main Results:

• MM-PBSA binding free energy demonstrated high convenience and success rates in predicting high-quality protein complex structures.
• Top-ranked poses achieved a 77% success rate for high-quality predictions and 90% for high- or medium-quality predictions.
• A scoring scheme based on computational alanine scanning mutagenesis yielded an 87% success rate for top-ranked high- or medium-quality poses.

Conclusions:

• MM-PBSA derived properties are highly effective in ranking protein:protein docking poses.
• The high ranking accuracy indicates the potential of these computational methods to predict near-native protein complex structures.
• This approach holds significant promise for advancing molecular biology research and accelerating drug discovery pipelines.