Related Experiment Videos
Rapid refinement of protein interfaces incorporating solvation: application to the docking problem
R M Jackson1, H A Gabb, M J Sternberg
1Biomolecular Modeling Laboratory, Imperial Cancer Research Fund, London, UK.
Journal of Molecular Biology
|March 26, 1998
Summary
This study introduces a computational method to refine protein-protein interfaces by modeling side-chain changes, solvation, and subunit movement. The approach improves the accuracy of protein docking, particularly for protease-inhibitor complexes, aiding in understanding protein interactions.
Area of Science:
- Computational Biology
- Structural Bioinformatics
- Protein Interaction Analysis
Background:
- Accurate modeling of protein-protein interactions is crucial for understanding biological processes.
- Existing protein docking methods often struggle to precisely refine interfaces, considering factors like side-chain flexibility and solvation.
Purpose of the Study:
- To develop a computationally efficient strategy for refining protein-protein interfaces.
- To model the effects of side-chain conformational changes, solvation, and limited rigid-body movements on protein complex stability.
- To assess the method's ability to discriminate between native and non-native protein-protein docked structures.
Main Methods:
- Proteins are represented at the atomic level with rotamer libraries for side-chains on a fixed peptide backbone.
- Solvation is modeled using Langevin dipoles, incorporating electrostatic, van der Waals, and hydrophobic interactions.
- A two-step energy refinement process involves iterative mean-field refinement of side-chain conformations followed by rigid-body minimization, repeated until convergence.
Main Results:
- The refinement method improved the root-mean-square deviation (RMSD) of side-chains in protease-inhibitor complexes, enhancing interface geometry.
- When solvation was included, the method successfully ranked native-like geometries within the top four solutions for four out of five protease-inhibitor complexes.
- Discrimination was less effective for antibody-antigen complexes, suggesting differences in binding mechanisms ('lock and key' vs. 'induced fit').
Conclusions:
- The developed computational strategy effectively refines protein-protein interfaces and shows promise in discriminating between native and non-native docked structures, especially for 'lock and key' binding types.
- The model's success with protease-inhibitors highlights the importance of side-chain flexibility and solvation in interface refinement.
- Further development to include more extensive conformational sampling could lead to a more generalized solution for the protein docking problem.