Data Integration Strategies in Metabolomics
Identification of metabolites is the central challenge for metabolomics. LC-MS has great sensitivity and is excellent for targeted studies or identification of unknowns when accurate database matching is possible. NMR has lower sensitivity than MS but offers outstanding reproducibility and the ability to obtain atomic-level structures necessary for unknown metabolite identification. Ultimately, integrating both techniques is ideal, because they often provide complementary information. For unknown molecules, both are needed for a confident identification. However, putting NMR and LC-MS datasets together remains a challenging problem. We present a method to integrate NMR, LC-MS, and quantitative flow cytometry data of Caenorhabditis elegans. An efficient MATLAB script allows any of these datasets to be correlated, linking molecular information with developmental stage. I will use several applications from our recent work to illustrate these approaches.