Related Experiment Videos
Automated analysis of protein NMR assignments using methods from artificial intelligence
D E Zimmerman1, C A Kulikowski, Y Huang
1Center for Advanced Biotechnology and Medicine and Department of Molecular Biology and Biochemistry, Rutgers University, Piscataway, NJ 08854-5638, USA.
Journal of Molecular Biology
|June 20, 1997
Summary
AUTOASSIGN, an expert system, accurately assigns protein resonances from NMR spectra. This automated approach achieves 98% sequence-specific assignments with minimal errors, accelerating structural biology research.
Area of Science:
- Biochemistry
- Structural Biology
- Computational Chemistry
Background:
- Nuclear Magnetic Resonance (NMR) spectroscopy is crucial for protein structure determination.
- Automating the complex process of resonance assignment in NMR spectra is a significant challenge.
Purpose of the Study:
- To develop and evaluate an expert system, AUTOASSIGN, for automated resonance assignment in protein NMR spectra.
- To assess the accuracy and efficiency of AUTOASSIGN across various protein sizes and spectral complexities.
Main Methods:
- Utilized a combination of symbolic constraint satisfaction and a domain-specific knowledge base.
- Integrated amino acid sequence with 2D (15N-1H) and 3D triple-resonance NMR spectral data.
- Applied the system to seven diverse protein datasets.
Main Results:
- Achieved an average of 98% sequence-specific spin-system assignments.
- Demonstrated a low error rate of less than 0.5% for assignments.
- Execution times ranged from seconds to minutes, showing computational efficiency.
Conclusions:
- AUTOASSIGN effectively automates protein resonance assignment from NMR data.
- The system's accuracy and speed facilitate structural biology workflows.
- Potential for extension to other NMR data types exists.