Improved NMR-data-compliant protein structure modeling captures context-dependent variations and expands the scope of...

Area of Science:

• Structural Biology
• Computational Biology
• Biophysics

Background:

• Nuclear magnetic resonance (NMR) spectroscopy is vital for determining protein structures and conformational states under physiological conditions.
• Large degrees of freedom and sparse NMR data can lead to structural artifacts in protein models.
• Current state-of-the-art methods often use molecular dynamics (MD) with simulated annealing for protein structure modeling.

Purpose of the Study:

• To develop an alternative graph-based modeling approach for protein structure determination from NMR data.
• To address limitations of existing methods in handling sparse data and large degrees of freedom.
• To generate enhanced NMR-evidence-based models that accurately reflect protein structure variations and support functional inference.

Main Methods:

• A graph-based modeling approach is proposed, building core substructures from NMR-derived distance-geometry constraints in one shot.
• A hybrid approach models regions with inadequate data while respecting distance-geometry constraints.
• The method is validated through detailed comparisons with state-of-the-art techniques and benchmarking across 106 protein folds.

Main Results:

• The novel approach successfully builds protein models with minimal experimental-constraint violations and adheres to standard structure quality criteria.
• Benchmarking on 106 protein folds (38-282 residues) demonstrates high accuracy and conformity to structural quality parameters.
• Comparative molecular dynamics (MD) studies reveal distinct conformational dynamics compared to state-of-the-art methods, potentially linked to protein function.

Conclusions:

• The developed graph-based method offers a robust alternative for building accurate protein models from NMR data, especially for systems with sparse information.
• The models generated capture secondary and tertiary structure variations, expanding the possibilities for functional inference.
• This approach enhances the utility of NMR data in understanding protein structure-function relationships.