Automated\nBeer Analysis by NMR Spectroscopy

Previously, we reported on the development\nof MagMet, a tool capable\nof automatically processing and quantifying 1D <sup>1</sup>H NMR spectra\nof complex chemical mixtures, including biofluids such as human serum\nor plasma and, more recently, beverages such as wine. In this article,\nwe present an extension of MagMet, called MagMet-B, for the automated\nprofiling of 1D <sup>1</sup>H NMR spectra of beer. We curated a comprehensive\n1D <sup>1</sup>H NMR spectral library comprising 81 more abundant\nmetabolites commonly found in beer samples and optimized the MagMet\nalgorithm to accurately fit these compounds. A comparison with manual\nprofiling using the Chenomx NMR Suite (Version 8.3) showed a strong\ncorrelation between the manually measured and automated MagMet metabolite\nconcentrations, with a mean absolute percent error of 13% and a median\nabsolute percent error of 9%. Time-to-process comparisons show that\nMagMet-B is up to 45× faster than manual analysis. The MagMet-B\nWeb server, which is specifically tailored for profiling beer NMR\nspectra at 700 MHz, is now accessible at https://magmet.ca.