Integrative Protein Modeling in RosettaNMR from Sparse Paramagnetic Restraints.

Area of Science:

• Biochemistry
• Structural Biology
• Computational Biology

Background:

• Predicting protein structure from sparse Nuclear Magnetic Resonance (NMR) data is challenging.
• Existing computational methods often struggle with limited experimental restraints, especially for larger proteins.
• Incorporating additional experimental data can enhance atomic detail in predicted structures.

Purpose of the Study:

• To introduce a comprehensive framework, RosettaNMR, within the Rosetta software suite for protein structure prediction.
• To integrate Nuclear Magnetic Resonance (NMR) restraints derived from paramagnetic labeling into computational structure prediction.
• To assess the performance of RosettaNMR in predicting structures of monomeric and oligomeric proteins and in docking applications.

Main Methods:

• Developed RosettaNMR, a framework incorporating pseudocontact shifts, residual dipolar couplings, and paramagnetic relaxation enhancements.
• Utilized existing Nuclear Magnetic Resonance (NMR) restraints including backbone chemical shifts and Nuclear Overhauser Effect (NOE) distance restraints.
• Validated RosettaNMR on 28 monomeric proteins, 8 homo-oligomeric proteins, 2 protein-protein docking, and 3 protein-ligand docking cases.

Main Results:

• Paramagnetic restraints significantly improved model accuracy for 85% of benchmark proteins.
• Combining paramagnetic restraints with chemical shifts yielded high-accuracy models (≤2Å) in 50% of cases.
• RosettaNMR demonstrated broad applicability in protein structure prediction and docking scenarios.

Conclusions:

• RosettaNMR provides a powerful computational approach for protein structure prediction using paramagnetic NMR restraints.
• The framework effectively addresses the challenge of sparse data by adding atomic detail.
• This method holds great potential for studying larger and more complex protein systems.