Nuclear Magnetic Resonance Spectroscopy Data Processing
In Chapter 3 an overview of NMR technology, how NMR data is produced by exploiting a phenomenon known as spin and the applications of NMR spectroscopy in metabolomics were presented. In this chapter, we focus on data—how we analyze NMR spectra to extract knowledge about metabolites—and we discuss data processing elements. We discuss what pre-processing and processing parameters can be used to improve the quality of a spectrum, how to best optimize these parameters and highlight potential pitfalls that can result from misuse of such parameters. A complete data processing pipeline (from raw data to identification) is examined. We compare fingerprinting, multi-integration of regions of interest (ROIs) and targeted profiling. We highlight the importance of region exclusion, normalization and scaling for robust chemometric analysis of NMR spectra. We also discuss different approaches for metabolite identification: the importance of online resources, statistical methods to aid identification and how the combination of NMR spectroscopy and mass spectrometry (MS) can be used to improve metabolite identification. We also demonstrate the application of NMR spectroscopy to Stable Isotope Resolved Metabolomics and structure elucidation and briefly introduce solid state Magic Angle Spinning (MAS) NMR for tissue metabolomics. We finish this chapter by giving an overview of what the future holds for NMR based metabolomics.