Public databases of NMR spectra of low molecular weight metabolites must be constructed to remove the major bottleneck of metabolite identification and quantification in the analysis of metabolomics data. Two-dimensional (2-D) 1H J-resolved spectroscopy represents a popular alternative to 1-D NMR methods, resolving the highly overlapped signals characteristic of complex metabolite mixtures across two frequency dimensions. Here we report the design, measurement and curation of, primarily, a database of 2-D J-resolved NMR spectra. Metabolites were selected based upon their importance within metabolic pathways and their detection potential by NMR, and prepared for analysis at pH 6.6, 7.0 and 7.4. Sixteen NMR spectra were recorded for each metabolite using a 500 MHz spectrometer, including 1-D and 2-D J-resolved spectra, different water suppression methods and different acquisition parameters. Some metabolites were removed due to limited solubility, poor NMR signal quality or contamination, and the final dataset comprised of 3328 NMR spectra arising from 208 metabolite standards. These data are housed in a purpose-built MySQL database (Birmingham Metabolite Library; BML-NMR) containing over 100 separate tables and allowing the efficient storage of raw free-induction-decays (FIDs), 1-D and 2-D NMR spectra and associated metadata. The database is compliant with the Metabolomics Standards Initiative (MSI) endorsed reporting requirements, with some necessary amendments. Library data can be accessed freely and searched through a custom written web interface (www. bml-nmr.org). FIDs, NMR spectra and associated metadata can be downloaded according to a newly implemented MSI-compatible XML schema.
Research Areas and Centers
- Academic Focus: Center for Infection and Inflammation Research (ZIEL)