Towards automated structure-based NMR resonance assignment

Richard Jang, Xin Gao, Ming Li*

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Scopus citations


We propose a general framework for solving the structure-based NMR backbone resonance assignment problem. The core is a novel 0-1 integer programming model that can start from a complete or partial assignment, generate multiple assignments, and model not only the assignment of spins to residues, but also pairwise dependencies consisting of pairs of spins to pairs of residues. It is still a challenge for automated resonance assignment systems to perform the assignment directly from spectra without any manual intervention. To test the feasibility of this for structure-based assignment, we integrated our system with our automated peak picking and sequence-based resonance assignment system to obtain an assignment for the protein TM1112 with 91% recall and 99% precision without manual intervention. Since using a known structure has the potential to allow one to use only N-labeled NMR data and avoid the added expense of using C-labeled data, we work towards the goal of automated structure-based assignment using only such labeled data. Our system reduced the assignment error of Xiong-Pandurangan-Bailey- Kellogg's contact replacement (CR) method, which to our knowledge is the most error-tolerant method for this problem, by 5 folds on average. By using an iterative algorithm, our system has the added capability of using the NOESY data to correct assignment errors due to errors in predicting the amino acid and secondary structure type of each spin system. On a publicly available data set for Ubiquitin, where the type prediction accuracy is 83%, we achieved 91% assignment accuracy, compared to the 59% accuracy that was obtained without correcting for typing errors.

Original languageEnglish (US)
Title of host publicationResearch in Computational Molecular Biology - 14th Annual International Conference, RECOMB 2010, Proceedings
Number of pages19
StatePublished - 2010
Event14th Annual International Conference on Research in Computational Molecular Biology, RECOMB 2010 - Lisbon, Portugal
Duration: Apr 25 2010Apr 28 2010

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume6044 LNBI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other14th Annual International Conference on Research in Computational Molecular Biology, RECOMB 2010

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'Towards automated structure-based NMR resonance assignment'. Together they form a unique fingerprint.

Cite this