Chemometric applications in Nuclear Magnetic Resonance metabonomics

The use of generalised 2D correlation spectroscopic (Gen2D) analysis techniques coupled with NMR spectroscopy has been proposed for identifying biomarkers and aiding in the modelling of biological processes. Spectral pre-processing is of paramount importance. During the analysis of dynamic 1H NMR spectra ‘ghost’ peaks can be observed in the two-dimensional asynchronous correlation plots. These have been demonstrated here, for the first time, to be the result of minor changes in peak widths between successive spectra. Such variation in spectra is common in NMR and can be removed if spectra are synthesized with constant peak widths.<br><br>Gen2D NMR correlation spectroscopy has been applied to monitoring a commercial fermentation of red wine. The synchronous matrix highlighted positively and negatively correlated spectral features and the asynchronous matrix provided information regarding the differing order of change of intensity of spectral features. These two matrices provide a diagrammatic summary of relative changes in samples over time. <br>Two-dimensional correlation methods have also been applied to interpretation of NMR-based spectra of plasma from human subjects undertaking an exercise-dietary regimen.<br><br>For the first time Gen2D, hybrid correlation analysis has been employed to monitor the metabolic effects of phosphonate in Phytophthora. In vivo 31P NMR was used to monitor changes in P-metabolites. <br><br>Finally, a multivariate analysis of plasma samples’ 1H- NMR spectra highlighted differences in the metabolic profiles of adult Black Bream fish exposed to a single high dose of the estrogen 17β-estradiol. <br><br>It is concluded that NMR 2D-correlation maps can be successful in aiding interpretation of time series of data, being visually accessible.