Related Experiment Video
Updated: Aug 19, 2025
Identification and Quantification of Deranged Metabolites in Critically Ill Patients Using NMR-Based Metabolomics
Published on: November 29, 2024
Problems, principles and progress in computational annotation of NMR metabolomics data.
Michael T Judge1, Timothy M D Ebbels2
1Section of Bioinformatics, Division of Systems Medicine, Department of Metabolism, Digestion and Reproduction, Imperial College, 131 Sir Alexander Fleming Building, South Kensington Campus, London, UK.
Automated tools can improve compound identification in Nuclear Magnetic Resonance (NMR) metabolomics by standardizing spectral matching and confidence scoring. This review aims to advance annotation standards and foster collaboration for better software solutions.
Area of Science:
- Metabolomics
- Analytical Chemistry
- Bioinformatics
Background:
- Compound identification is a major challenge in Nuclear Magnetic Resonance (NMR) metabolomics, particularly with 1-dimensional proton (1H 1D) data.
- Current workflows heavily depend on database searches for metabolite identification, but validation and communication of annotations lack standardized practices.
- Expert knowledge guides annotation validation, leading to variability and hindering community-wide standardization efforts.
Purpose of the Study:
- To broaden the use of automated annotation tools in NMR metabolomics.
- To establish standardized terminology for classifying spectral matching information and communicating annotation confidence.
- To encourage collaboration between data scientists, software developers, and the NMR metabolomics community.
Main Methods:
- Discusses the typical untargeted NMR identification workflow, distinguishing between annotation and identification.
- Examines three key aspects of annotation: query generation, spectral matching against reference data, and scoring/confidence estimation.
- Highlights existing automated and semi-automated annotation approaches based on utilized structural information and its computational representation.
Main Results:
- The review differentiates annotation (hypothesis generation) from identification (hypothesis testing).
- It details the utility of various NMR data features for annotation.
- It explores computational representations of structural information for spectral matching.
Conclusions:
- Standardizing spectral matching and confidence estimation is crucial for reliable compound identification in NMR metabolomics.
- Developing clear terminology and robust automated tools will enhance the reproducibility and communication of metabolomic data.
- Fostering interdisciplinary collaboration is key to advancing NMR metabolomics software and analytical standards.
