NMR-assisted protein structure prediction with MELDxMD

Area of Science:

• Biochemistry
• Structural Biology
• Computational Biology

Background:

• Protein structure determination is crucial for understanding biological function.
• Nuclear Magnetic Resonance (NMR) spectroscopy provides valuable experimental data for structural studies.
• Traditional methods can be limited by the complexity and size of protein targets.

Purpose of the Study:

• To evaluate the performance of MELD-accelerated molecular dynamics (MELDxMD) for protein structure determination using NMR data.
• To assess MELDxMD's effectiveness in the NMR-assisted category of the Critical Assessment of protein Structure Prediction (CASP) competition.

Main Methods:

• MELDxMD simulations were initiated using web server predictions.
• The method utilized NMR-derived data, including NOE-based contacts and backbone dihedrals.
• Simulations were performed on various protein targets, including a large 326-mer protein.

Main Results:

• MELDxMD achieved the best performance in the NMR-data-assisted category of CASP13 across 17 targets.
• The method outperformed other groups by approximately a factor of 4 in z-score.
• Predicted structures showed good agreement with experimental residual dipolar couplings.

Conclusions:

• MELDxMD is a highly effective method for protein structure determination from NMR data.
• The approach can successfully integrate noisy and ambiguous experimental information to refine molecular dynamics searches.
• MELDxMD shows significant promise for advancing structural biology research using NMR constraints.