Accurate quantification of volatile fatty acid (VFA) concentrations in rumen fluid are essential for research on rumen metabolism. The study comprehensively investigated the pros and cons of High-performance liquid chromatography (HPLC) and 1H Nuclear magnetic resonance (1H-NMR) analysis methods for rumen VFAs quantification. We also investigated the performance of several commonly used data pre-treatments for the two sets of data using correlation analysis, principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA). The molar proportion and reliability analysis demonstrated that the two approaches produce highly consistent VFA concentrations. In the pre-processing of NMR spectra, line broadening and shim correction may reduce estimated concentrations of metabolites. We observed differences in results using multiplet of different protons from one compound and identified "handle signals" that provided the most consistent concentrations. Different data pre-treatment strategies tested with both HPLC and NMR significantly affected the results of downstream data analysis. "Normalized by sum" pre-treatment can eliminate a large number of positive correlations between NMR-based VFA. A "Combine" strategy should be the first choice when calculating the correlation between metabolites or between samples. The PCA and PLS-DA suggest that except for "Normalize by sum", pre-treatments should be used with caution.
Bibliographical noteFunding Information:
The authors acknowledge John Rooke for his contributions in HPLC analysis and Riccardo Bica for his contributions in NMR analysis. This research is jointly supported by Ulster University and Scotland’s Rural College, U.K.
The research was completed as part of MY’s Ph.D. which is jointly funded by Ulster University and Scotland’s Rural College, U.K.
© 2021, The Author(s).
- High-performance liquid chromatography
- Nuclear magnetic resonance
- Data pre-processing
- Volatile Fatty Acid