Structure prediction of protein complexes by an NMR-based protein docking algorithm

Area of Science:

• Structural biology
• Computational chemistry
• Biophysics

Background:

• Protein docking algorithms predict complex structures, aiding experimental elucidation.
• Integrating experimental data like NMR spectra can refine docking predictions.

Purpose of the Study:

• To develop and evaluate a new scoring function for protein-protein docking using 1H-NMR spectra.
• To assess the function's ability to rank potential complex structures accurately.

Main Methods:

• A novel scoring function was developed, comparing theoretical 1H-NMR spectra of docked complexes with experimental data.
• The function ranks structures based on the difference spectra, avoiding distance constraints.
• The approach was tested on four protein complexes using PDB structures and BMRB NMR assignments.

Main Results:

• The new scoring function successfully ranked potential complex structures.
• It demonstrated superior performance compared to standard scoring functions, particularly in challenging cases.
• Achieved near-perfect separation between correct and incorrect complex structures in one instance where others failed.

Conclusions:

• The proposed NMR-based scoring function is a powerful tool for evaluating protein-protein docking models.
• This method offers an alternative to traditional approaches, especially when standard scoring functions are insufficient.
• The findings suggest a promising direction for integrating spectroscopic data into computational structural biology.