Recent Advances in Targeted and Untargeted Metabolomics by NMR and MS/NMR Methods

Abstract

Metabolomics has made significant progress in multiple fronts in the last 18 months. This minireview aimed to give an overview of these advancements in the light of their contribution to targeted and untargeted metabolomics. New computational approaches have emerged to overcome the manual absolute quantitation step of metabolites in one-dimensional (1D) 1H nuclear magnetic resonance (NMR) spectra. This provides more consistency between inter-laboratory comparisons. Integration of two-dimensional (2D) NMR metabolomics databases under a unified web server allowed for very accurate identification of the metabolites that have been catalogued in these databases. For the remaining uncatalogued and unknown metabolites, new cheminformatics approaches have been developed by combining NMR and mass spectrometry (MS). These hybrid MS/NMR approaches accelerated the identification of unknowns in untargeted studies, and now they are allowing for profiling ever larger number of metabolites in application studies.

Metabolomics in chronic lung diseases

“…Metabolomics is generally applied through a targeted, selective measurement strategy, or as part of an untargeted, global profiling approach . Targeted metabolomics is a quantitative approach that allows for the measurement of metabolite concentrations while untargeted metabolomics undertakes simultaneous assessment of metabolites without any prior sample knowledge for hypothesis generation . Untargeted metabolomics is a comprehensive strategy for identifying changes in different pathophysiological states; however, a major disadvantage is that the majority of features are unidentifiable.…”

“…Untargeted metabolomics is a comprehensive strategy for identifying changes in different pathophysiological states; however, a major disadvantage is that the majority of features are unidentifiable. Feature identification can be improved by using sophisticated data‐independent acquisition techniques, which will additionally collect important structural detail in the metabolite measurements for interpretation at a later date without the need for re‐measurement of samples as a single mass measurement alone is insufficient for compound identification . Additional orthogonal information, such as chromatographic retention time, tandem mass spectrometry fragmentation (MS/MS), isotopic patterns and collisional cross‐section, is also necessary for structural annotation and compound identification .…”

“…Feature identification can be improved by using sophisticated data‐independent acquisition techniques, which will additionally collect important structural detail in the metabolite measurements for interpretation at a later date without the need for re‐measurement of samples as a single mass measurement alone is insufficient for compound identification . Additional orthogonal information, such as chromatographic retention time, tandem mass spectrometry fragmentation (MS/MS), isotopic patterns and collisional cross‐section, is also necessary for structural annotation and compound identification . Advents in analytical technologies and bioinformatics are setting the stage for large‐scale, harmonized metabolic profiling suitable for studies at the population level .…”

A Perspective and Framework for Developing Sample Type Specific Databases for LC/MS-Based Clinical Metabolomics

“…Small molecules/compounds often share identical masses, making it challenging to accurately identify compounds based solely on mass and/or molecular formula [4,5]. While other technologies, such as gas chromatography mass spectrometry (GC/MS) and nuclear magnetic resonance (NMR), address many of these issues [6,7] through, for example, comprehensive GC/MS spectral libraries, many metabolomics projects are proposed sample type specific database (STSDB) searching workflows. The traditional workflow for annotating metabolomics data using databases is shown on the left.…”

ASICS: an R package for a whole analysis workflow of 1D 1H NMR spectra

“…Not only is this identification step tedious, time consuming, expert dependent 19 and not reproducible but it also leads to a serious loss of information since the identi-20 fication of metabolites is restricted to the ones that correspond to extracted buckets 21 (Considine et al, 2018). 22 Some methods have thus been developed to automatically identify metabolites from 23 1 H NMR spectra (MetaboHunter (Tulpan et al, 2011), MIDTool (Filntisi et al, 2017)) 24 and others to automatically quantify the concentration of detected metabolites (Autofit 25 (Weljie et al, 2006), batman (Hao et al, 2012), Bayesil (Ravanbakhsh et al, 2015) and 26 rDolphin (Cañueto et al, 2018)); see Bingol (2018) for a complete review. Recently, 27 Tardivel et al (2017) defined a new statistical method to automatically identify and 28 quantify metabolites that outperforms the other approaches.…”