NASCA

NASCA automates side-chain resonance and NOE assignments from NMR data to facilitate protein structure determination.

Key Features:

• Markov Random Field (MRF) formulation: NASCA formulates the assignment problem as a Markov Random Field (MRF).
• Combinatorial protein design algorithms: It leverages combinatorial protein design algorithms to compute optimal assignments under the MRF.
• NOESY-derived contact maps: The MRF integrates contact map information derived from NOESY spectra.
• RDC integration: The MRF incorporates backbone structural insights obtained from residual dipolar couplings (RDCs).
• Side-chain rotamer consideration: NASCA considers all possible side-chain rotamers during assignment evaluation.
• Dead-end elimination (DEE): It employs dead-end elimination to prune non-optimal side-chain resonance assignments.
• A* search: An A* search algorithm is used to identify a set of optimal side-chain resonance assignments.
• Avoidance of through-bond experiments: NASCA eliminates the need for through-bond experiments such as HCCH-TOCSY or HCCCONH.
• NOE distance restraints output: The computed assignments are used to generate NOE distance restraints for structure calculation.
• Empirical performance: On five proteins NASCA achieved >90% side-chain proton assignment with ~80% correct assignment accuracy and produced structures with backbone RMSD of 0.8–1.5 Å to reference NMR structures.

Scientific Applications:

• Automated NMR assignment: Automating side-chain resonance and NOE assignments in protein NMR spectroscopy.
• NOE ambiguity resolution: Resolving NOE assignment ambiguities to improve restraint quality.
• Protein structure determination: Enabling high-resolution protein structure determination from NOE distance restraints.
• Large-protein applicability: Applicable to larger proteins where through-bond experiments like HCCH-TOCSY or HCCCONH are less effective.
• Integrated NOESY and RDC interpretation: Integrating NOESY-derived contact maps with RDC-derived backbone information to interpret NMR data.

Methodology:

Formulate assignments as a Markov Random Field integrating NOESY-derived contact maps and RDC backbone information; evaluate assignments over side-chain rotamers using combinatorial protein design algorithms; prune with dead-end elimination (DEE); and search optimal assignments with an A* algorithm to produce NOE distance restraints.