Abstract

Identification of metabolites and inference of differentially regulated metabolic pathways are the key aims of metabolomics research. A common approach to this problem is to compare the experimental spectrum with the database archived metabolite information. However, databases store metabolite information at “standard condition” which are in most cases in variance to the experimental condition data is obtained. Consequently, a considerable fraction of the metabolites remain unidentified and the experimental spectrum unharvested.

We have developed a method based on matching the pattern of spectrum peaks rather than absolute tolerance thresholds, using a combination of geometric hashing and similarity scoring techniques. When applied to 2D NMR metabolomics data, tests with 719 metabolites from the Human Metabolome Database show that 100% of the metabolites can be assigned correctly when accurate data are available. A high success rate is obtained even in the presence of large chemical shift deviations such as 0.5 ppm in ¹H and 3 ppm in ¹³C and missing peaks (up to 50%), compared to nearly no assignments obtained under these conditions with existing methods that employ a direct database search approach. A variation of the approach has been extended to obtain “peaks to pathways” information from NMR spectral data.