Heuristic search for 2D NMR alignment to support metabolite identification

  • Authors:

  • Geun-Cheol Lee;Jeff de Ropp;Mark R. Viant;David L. Woodruff;Ping Yu

  • Affiliations:

  • Konkuk University/ Seoul/ Korea;University of California, Davis, Davis CA;The University of Birmingham, Birmingham, UK;University of California, Davis, Davis CA;University of California, Davis, Davis CA

  • Venue:
  • ESCAPE'07 Proceedings of the First international conference on Combinatorics, Algorithms, Probabilistic and Experimental Methodologies
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

For the problem of aligning two-dimensional NMR spectra of biological samples to determine if metabolite standards in a database can be excluded as possible constituents, we develop heuristic search algorithms that offer tremendous time savings when compared to manual methods. Xi et al [15] consider this problem and use statistical methods to reduce the search space and enumerate it. In this paper we consider the case when the statistical model is not available due to lack of data. We describe a simulated annealing algorithm and an algorithm that hybridizes simulated annealing with a shift neighborhood and a variant of reactive tabu search with a large neighborhood. Computational experiments based on data from physical experiments demonstrates that the hybrid is more effective than its constituents for runs with limited CPU time but that simulated annealing and the hybrid are roughly equal for longer runs.