Modeling side-chains using molecular dynamics improve recognition of binding region in CAPRI targets

Area of Science:

• Computational Biology
• Structural Biology
• Biophysics

Background:

• Predicting protein-protein interactions is crucial for understanding biological processes.
• The CAPRI-II experiment presented challenges in predicting 3D protein structures, including receptor or ligand.
• Accurate modeling of side-chain conformations is vital for precise molecular recognition.

Purpose of the Study:

• To enhance protein-protein interaction prediction by integrating molecular dynamics (MD) with computational docking.
• To assess the accuracy of MD in predicting side-chain rotamer conformations for molecular recognition.
• To evaluate a free energy-based scoring function for discriminating native-like protein complex structures.

Main Methods:

• Utilized molecular dynamics (MD) simulations in explicit solvent to model key side-chain conformations.
• Employed rigid-body docking server ClusPro and flexible docking algorithm SmoothDock.
• Applied a free energy scoring function combining van der Waals, electrostatic, and desolvation energies.

Main Results:

• MD simulations accurately predicted rotamer conformations of anchor side-chains critical for binding.
• Conformations predicted by MD were often common rotamers, while induced-fit side-chains adopted less common ones.
• The integrated approach yielded successful predictions for 6 out of 9 CAPRI-II targets, excluding those with significant structural rearrangements.

Conclusions:

• Modeling side-chain dynamics with MD significantly improves the accuracy of computational docking for protein-protein interactions.
• The free energy-based scoring function effectively identifies native-like conformations, highlighting the role of thermodynamics.
• The methodology shows promise for predicting protein interactions, particularly for systems without major conformational changes upon binding.