Issue No. 01 - January/February (2002 vol. 4)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/5992.976437
<p>NMR resonance assignment is one of the key steps in solving an NMR protein structure. The assignment process is tedious and time-consuming, and although many computer programs can assist the assignment process, several NMR labs still do the assignments manually or semimanually for quality reasons. The authors present a new computational framework for automating the assignment process, particularly for backbone resonance peak assignment. They formulate the assignment problem as a constrained weighted bipartite matching problem, which provides a natural framework for incorporating all available information into the assignment process. Although they can show that the constrained bipartite matching problem, in the most general situation, is NP-hard, they developed a rigorous algorithm for solving the problem. For practical data, the algorithm runs fast, taking advantage of the discerning power of their scoring scheme and ruling out a vast majority of bad individual assignments in a preprocessing step. They have implemented the algorithm and tested it on four proteins with both real and simulated NMR peaks. Their results show that when 75 percent of all the connectivities between NMR peaks are identified with simulated data, their assignment results will be close to or reach 100 percent accuracy. With the realistic connectivity information obtained from experimental peaks, their program can assign majority of the peaks correctly.</p>
V. Olman, T. Jiang, Y. Xu, D. Xu, D. Kim and J. Razumovskaya, "Automated Assignment of Backbone NMR Peaks Using Constrained Bipartite Matching," in Computing in Science & Engineering, vol. 4, no. , pp. 50-62, 2002.