Abstract
Metabonomic analysis of urine utilizing high-resolution NMR spectroscopy and chemometric techniques has proven valuable in characterizing the biochemical response to an intervention. To assess the effect of magnetic field strength on information contained in NMR-based metabonomic data sets, 1H NMR spectra were acquired on 250-, 400-, 500-, and 800-MHz instruments, respectively, on the same set of human urine samples collected before and after dietary interventions with milk and with meat proteins. Partial least-squares regression discriminant analyses (PLS-DA) were performed in order to elucidate the ability of the 1H spectra acquired at various field strengths to identify possible spectral differences and discriminate between pre- and postintervention samples. The loadings from PLS-DA contained the same spectral regions, implying that the same metabolites were involved in the discrimination independent of magnetic field strength. The investigation revealed a strong increase in prediction performance and thereby spectral information content when increasing the magnetic field strength from 250 to 500 MHz, while from 500 to 800 MHz the increase was less pronounced.
Originalsprog | Engelsk |
---|---|
Tidsskrift | Analytical Chemistry |
Vol/bind | 79 |
Udgave nummer | 18 |
Sider (fra-til) | 7110-7115 |
Antal sider | 6 |
ISSN | 0003-2700 |
DOI | |
Status | Udgivet - 2007 |