Skip to content
Education
Isle of NAM
Metabolomics

Metabolomics

Table of Contents

Definition

Metabolomics is the systematic and comprehensive qualitative and quantitative characterization of the metabolome, the complete complement of low-molecular-weight metabolites present within a biological system, encompassing endogenous intermediates and end products of metabolism as well as compounds of dietary, microbial, and xenobiotic origin1 2. Metabolites, the small molecules that are transformed during metabolism, provide a direct functional readout of cellular biochemistry, and their profiling at the systems level can be correlated with phenotype and used to identify disease mechanisms and therapeutic targets1.

The biological basis of metabolomics rests on the organization of metabolism into enzyme-catalyzed reaction networks that convert nutrients into energy and biomass. Metabolite pool sizes are governed by the integrated activity of transporters, enzymes, and regulatory proteins across central carbon and intermediary metabolism, including glycolysis, the tricarboxylic acid cycle, amino-acid biosynthesis, and lipid metabolism1. Because metabolite concentrations respond simultaneously to gene expression, enzyme activity, host microbiota, diet, drugs, and environmental exposures, the metabolome is positioned at the terminal layer of the omics cascade, downstream of the genome, transcriptome, and proteome, and models the functional metabolic phenotype of a cell, tissue, organ, or whole organism1 3. Analytically, the metabolome is sampled from cells, tissues, and accessible biofluids such as blood plasma, serum, and urine, in which metabolites serve as reporters of systemic biochemical state2.

Purpose

The primary value proposition of metabolomics is that metabolite levels constitute the measurement layer closest to the observable phenotype. Whereas genomic and transcriptomic measurements describe the informational potential of a system, metabolite concentrations represent the integrated output of cellular regulation, encoding the functional consequences of genetic variants, environmental exposures, and disease processes1 3. A single metabolic profile simultaneously reports on hundreds of biochemical pathways, capturing pathway-level perturbations that no single-analyte measurement can reveal1 3.

Metabolomics improves upon legacy models of biochemical measurement in two principal respects. First, whereas conventional clinical chemistry and hypothesis-driven analytical biochemistry quantify one or a small number of analytes selected a priori, metabolomics can measure hundreds to thousands of metabolites in a single analysis, permitting the discovery of biomarkers and pathway perturbations that were not anticipated by the study hypothesis1 4. Second, whereas genetic risk models describe inherited probability, the metabolome integrates genetic predisposition with environmental and lifestyle exposures, thereby reporting the actual current biochemical state of the individual3.

The human relevance of the technology is direct. Metabolomics is performed on human biofluids and tissues obtained by routine, minimally invasive collection, and the human metabolome itself has been catalogued in more than 114,000 curated entries2. Metabolic biomarkers are under active investigation for the early detection and monitoring of cancer, cardiovascular disease, metabolic syndrome, and neurodegenerative disorders4, and deep metabolic phenotyping is positioned as a foundation for precision medicine and patient stratification3.

Methodology

A metabolomics experiment proceeds through a defined workflow: study design and pre-analytical control of sample collection; quenching of metabolism and solvent-based extraction of metabolites, with chemical derivatization typically required to render analytes volatile for gas chromatography; chromatographic separation and detection; computational processing of raw spectra; statistical analysis; and metabolite annotation against reference libraries1 5.

Three analytical platforms dominate the field: liquid chromatography coupled to mass spectrometry (LC-MS), gas chromatography coupled to mass spectrometry (GC-MS), and nuclear magnetic resonance (NMR) spectroscopy5. LC-MS and GC-MS separate metabolite mixtures chromatographically, ionize them by electrospray or electron impact, and measure mass-to-charge ratios with high-resolution analyzers; tandem mass spectrometry (MS/MS) provides fragmentation patterns used for structural elucidation5 6. NMR spectroscopy measures the magnetic resonance of nuclei such as 1H, 13C, and 31P, is nondestructive, requires minimal sample preparation, supports absolute quantification, and exhibits high experimental reproducibility, which makes it the preferred platform for long-term and large-scale clinical studies5.

Two complementary experimental designs are employed. Targeted metabolomics quantifies a predefined panel of known metabolites, typically on triple-quadrupole instruments operated in multiple reaction monitoring mode, with authentic standards and stable-isotope-labeled internal standards enabling absolute quantification6. Untargeted metabolomics instead profiles as many spectral features as possible without prior analyte selection, producing relative abundance data of maximal biochemical breadth and serving as the hypothesis-generating arm of the field1 7.

Metabolite identification is performed by matching accurate mass, retention behavior, and MS/MS spectra against reference libraries such as the Human Metabolome Database and community spectral repositories, with the confidence of each identification graded according to the confidence-level scheme of the Metabolomics Standards Initiative2 7. Quality assurance relies on pooled quality-control samples, blanks, and reference materials, notably the NIST Standard Reference Material 1950 human plasma, which is widely used to validate methods and to monitor analytical variance across laboratories and studies8. Processed data are analyzed with univariate and multivariate chemometric methods, and completed datasets are deposited in public repositories such as Metabolomics Workbench, which provides standardized metadata, metabolite standards, protocols, and analysis tools for the community7 9.

Performance

Empirical coverage. No single analytical platform can completely identify and quantify all metabolites in a typical biological sample, and the most rigorous studies therefore employ multiple platforms5. The Human Metabolome Database, the most comprehensive organism-specific metabolite resource, catalogues more than 114,000 metabolite entries for humans, of which a substantial fraction are expected or predicted rather than experimentally detected compounds2. Critical analyses of chemical databases further conclude that existing resources cannot comprehensively retrieve even the known metabolites of a species, so the true size of any metabolome remains experimentally undetermined10.

Annotation performance. In untargeted mass-spectrometric profiling, on average only approximately ten percent of detected molecules can be annotated, and this low annotation rate constrains biochemical interpretation and the comparison of results between studies7.

Reproducibility. In the multi-instrument metabo-ring ring trial, which distributed a common set of samples to five NMR instruments and eleven mass spectrometers across laboratories, a high convergence in the spectral information describing the same samples was observed irrespective of standardization approach, deconvolution method, or instrument configuration, supporting the reliability of untargeted profiling at the level of sample characterization11. NMR exhibits the highest inter-measurement reproducibility among the principal platforms and scales robustly to large clinical series5. Targeted platforms quantify metabolites absolutely with isotope-dilution internal standards, achieving specificity and sensitivity sufficient for clinical deployment6.

Operational scale. The maturity of the technology is evidenced by population-scale deployment: 249 quantitative NMR biomarkers have been measured in plasma from 118,461 participants of the UK Biobank, demonstrating that standardized metabolomics can operate at the scale of national biobank cohorts12.

Technical boundaries. The principal limitations of the field are the annotation bottleneck described above7; residual pre-analytical and inter-laboratory variance, which motivates the continued development of reference materials and quality-control frameworks8; and incomplete clinical translation, in which targeted panels are established in routine care for specific indications while untargeted profiling remains largely investigational, with reported diagnostic sensitivities of approximately 87 percent in pilot series for inborn errors of metabolism13.

Applications

Clinical diagnostics. Targeted metabolomics is the analytical foundation of newborn screening and of the biochemical diagnosis and monitoring of inborn errors of metabolism, in which predefined metabolite panels have been used for decades; untargeted metabolomics is emerging as a next-generation screening modality capable of revealing disorders not covered by fixed panels13.

Biomarker discovery. Metabolomic profiling is extensively applied to the discovery and validation of biomarkers for diagnosis, prognosis, and treatment monitoring in cancer, cardiovascular disease, metabolic syndrome, and neurodegenerative disorders, where metabolic signatures enable patient stratification and early detection4.

Drug development and toxicology. Metabolomics is incorporated into preclinical drug safety assessment, where it elucidates adverse outcome pathways by detecting endogenous biochemical alterations underlying toxicity mechanisms, informs routine safety screening, and supports regulatory decision-making14.

Population health. NMR-based metabolic phenotyping operates at national biobank scale, supporting disease-wide association analyses of 249 circulating biomarkers and the translation of metabolic epidemiology into public health research12.

Precision medicine. Deep metabolic phenotyping of individuals enables pharmacometabolomics, the characterization of disease heterogeneity and progression, and the identification of biomarkers and targets for clinical intervention, positioning metabolomics as a core technology for precision medicine3.

Deployment stage and readiness. Targeted assays are deployed in routine clinical practice; NMR biobank phenotyping is operational at population scale; and untargeted workflows remain primarily research-stage, progressing toward clinical readiness through standardized reporting, quality-control frameworks, and open data deposition in repositories such as Metabolomics Workbench9 12 13.

Variations

Targeted Metabolomics — The quantitative analysis of predefined metabolite panels, typically by LC-MS/MS in multiple reaction monitoring mode with authentic and stable-isotope-labeled standards; the design employed in clinical research, newborn screening, and validated diagnostic assays6.

Untargeted Metabolomics — The global profiling of as many spectral features as possible without prior analyte selection, prioritizing biochemical breadth and discovery over absolute quantification; constrained by low annotation rates and relative rather than absolute measurements1 7.

Lipidomics — The branch of metabolomics dedicated to the large-scale study of cellular lipids and their pathways and networks, employing LC-MS, GC-MS, direct-infusion MS, and imaging MS approaches, with applications in diabetes, cardiovascular disease, and biomarker research15.

Fluxomics — Stable-isotope tracing, in which labeled substrates such as 13C-glucose are administered and isotopic labeling patterns are measured to quantify metabolic flux rates, providing dynamic information that static metabolite concentrations cannot supply16.

Spatial Metabolomics — Imaging mass spectrometry, including MALDI and DESI approaches, that maps the spatial distribution of metabolites across tissue sections, preserving anatomical context that is lost in bulk extraction workflows17.

  1. Patti GJ, Yanes O, Siuzdak G, Innovation: Metabolomics: the apogee of the omics trilogy, Nature Reviews Molecular Cell Biology, 22 March 2012.
    This review positions metabolomics as the terminal layer of the omics cascade, arguing that metabolites provide a direct functional readout of cellular biochemistry that can be correlated with phenotype and used to identify therapeutic targets. It surveys emerging mass spectrometric and bioinformatic technologies, their role in diagnostics and cell-metabolism research, and the challenges that remain for the field. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  2. Wishart DS, Guo AC, Oler E, et al., HMDB 5.0: the Human Metabolome Database for 2022, Nucleic Acids Research, 7 January 2022.
    This database paper documents HMDB 5.0, the world’s largest and most comprehensive organism-specific metabolomic resource, containing more than 114,000 richly annotated metabolite entries with chemical, clinical, and biochemical information. It describes upgrades to the database interface and content that support applications in human metabolomics, exposomics, lipidomics, nutritional science, and clinical chemistry. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  3. Beger RD, Dunn WB, Bandukwala A, et al., Metabolomics enables precision medicine: “A White Paper, Community Perspective”, Metabolomics, 2 September 2016.
    This community white paper articulates how deep metabolic phenotyping enables precision medicine, providing a readout of the functions of genetic variants, mechanistic insight into disease pathways, heterogeneity, progression, and treatment response. It reviews pharmacometabolomics and defines the practical requirements for translating metabolomics into clinical practice. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  4. Vo D-K, Trinh KTL, et al., Emerging Biomarkers in Metabolomics: Advancements in Precision Health and Disease Diagnosis, International Journal of Molecular Sciences, 8 December 2024.
    This review surveys recent metabolomics-based biomarker discovery and its transformative impact on precision health, covering oncometabolites in cancer, early-detection and treatment-response biomarkers, and signatures in neurodegenerative, cardiovascular, and metabolic disorders. It also discusses how metabolic biomarkers support patient stratification and personalized treatment. ↩︎ ↩︎ ↩︎

  5. Emwas A-H, Roy R, McKay RT, et al., NMR Spectroscopy for Metabolomics Research, Metabolites, 27 June 2019.
    This review describes NMR spectroscopy as one of the three principal analytical platforms of metabolomics alongside GC-MS and LC-MS, detailing its minimal sample preparation, absolute quantification, high experimental reproducibility, and nondestructive nature. It explains why no single platform can cover the entire metabolome and why NMR is preferred for long-term and large-scale clinical studies. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  6. Anh NK, Thu NQ, Tien NTN, Long NP, Nguyen HT, Advancements in Mass Spectrometry-Based Targeted Metabolomics and Lipidomics: Implications for Clinical Research, Molecules, December 2024.
    This review summarizes recent advances in targeted mass-spectrometry metabolomics and lipidomics for clinical research, including microsampling, dynamic multiple reaction monitoring, and ion-mobility mass spectrometry. It highlights the quantitative strengths of targeted designs for diagnosis and precision treatment and the challenges that persist in sample collection, quantification, and quality control. ↩︎ ↩︎ ↩︎ ↩︎

  7. de Jonge NF, Mildau K, Meijer D, et al., Good practices and recommendations for using and benchmarking computational metabolomics metabolite annotation tools, Metabolomics, 5 December 2022.
    This review establishes good practices for computational metabolite annotation, noting that in untargeted mass spectrometry on average only ten percent of detected molecules can be annotated. It benchmarks spectral-library and machine-learning annotation tools, discusses Metabolomics Standards Initiative identification confidence levels, and frames annotation as the principal bottleneck of the field. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  8. Lippa KA, Aristizabal-Henao JJ, Beger RD, et al., Reference materials for MS-based untargeted metabolomics and lipidomics: a review by the metabolomics quality assurance and quality control consortium (mQACC), Metabolomics, 9 April 2022.
    This consortium review catalogs the reference materials used for quality assurance and quality control in untargeted metabolomics and lipidomics, including pooled plasma materials such as NIST SRM 1950. It explains how the routine use of reference materials improves comparability of results across laboratories and studies and identifies standardization gaps that the community must close. ↩︎ ↩︎

  9. Sud M, Fahy E, Cotter D, et al., Metabolomics Workbench: An international repository for metabolomics data and metadata, metabolite standards, protocols, tutorials and training, and analysis tools, Nucleic Acids Research, 4 January 2016.
    This resource paper describes Metabolomics Workbench, a public international repository for metabolomics experimental data and standardized metadata spanning multiple species and analytical platforms. It documents the repository’s metabolite standards, structures, protocols, tutorials, and analysis tools that support transparent data sharing and reanalysis. ↩︎ ↩︎

  10. Kind T, Scholz M, Fiehn O, How Large Is the Metabolome? A Critical Analysis of Data Exchange Practices in Chemistry, PLOS ONE, May 2009.
    This analysis examines whether the size of a species’ metabolome can be determined, concluding that genome-guided reconstruction misses the products of orphan genes and that existing chemical databases cannot comprehensively retrieve all known metabolites. It argues that metabolomes must be determined experimentally and calls for machine-readable data exchange practices in chemistry. ↩︎

  11. Martin J-C, Maillot M, Mazerolles G, et al., Can we trust untargeted metabolomics? Results of the metabo-ring initiative, a large-scale, multi-instrument inter-laboratory study, Metabolomics, 2015.
    This large-scale ring trial distributed a common set of human samples to five NMR instruments and eleven mass spectrometers across laboratories to assess the reliability of untargeted metabolomics. Its main finding is a high convergence in the spectral information produced by the various instruments and configurations, supporting the reliability of untargeted profiling at the level of sample characterization. ↩︎

  12. Julkunen H, Cichonska A, Tiainen M, et al., Atlas of plasma NMR biomarkers for health and disease in 118,461 individuals from the UK Biobank, Nature Communications, 1 February 2023.
    This study reports plasma NMR biomarker data covering 249 measures of lipoprotein lipids, fatty acids, amino acids, ketones, and glycolysis metabolites for 118,461 UK Biobank participants. It demonstrates that standardized NMR metabolomics can operate at population scale and reveals systematic associations between circulating biomarkers and a broad spectrum of diseases. ↩︎ ↩︎ ↩︎

  13. Hertzog A, Selvanathan A, et al., A narrative review of metabolomics in the era of “-omics”: integration into clinical practice for inborn errors of metabolism, Translational Pediatrics, October 2022.
    This narrative review examines the integration of metabolomics into clinical practice for inborn errors of metabolism, contrasting decades-old targeted diagnostic panels with emerging untargeted next-generation screening. It reports diagnostic sensitivities on the order of 87 percent in pilot series for untargeted screening and discusses how metabolic derangements can assist molecular diagnostics. ↩︎ ↩︎ ↩︎

  14. Sille F, et al., Metabolomics in Preclinical Drug Safety Assessment: Current Status and Future Trends, Metabolites, 31 January 2024.
    This review assesses the current status and future trends of metabolomics in preclinical drug safety assessment, showing how endogenous biochemical alterations reveal adverse outcome pathways underlying toxicity mechanisms. It discusses the role of metabolomics in safety screening, mechanistic toxicology, and regulatory decision-making, and its integration with computational and in vitro innovations. ↩︎

  15. Wu Z, Shon JC, Mass Spectrometry-based Lipidomics and Its Application to Biomedical Research, Journal of Lifestyle Medicine, 31 March 2014.
    This review defines lipidomics as the large-scale study of the pathways and networks of all cellular lipids and compares four representative profiling technologies: GC-MS, LC-MS, direct-infusion MS, and imaging MS. It summarizes lipid databases and discusses applications of lipidomics to the diagnostics of diseases such as diabetes. ↩︎

  16. Kim I-Y, Park S, Kim Y, Kim H-J, Wolfe RR, Tracing metabolic flux in vivo: basic model structures of tracer methodology, Experimental & Molecular Medicine, September 2022.
    This review presents the basic model structures of stable-isotope tracer methodology, showing that flux calculations rest on four fundamental structures: tracer dilution in single- or multiple-pool systems and tracer incorporation from single or multiple precursors in steady or nonsteady states. It demonstrates how isotope tracing quantifies dynamic metabolic fluxes that static omics measurements cannot capture. ↩︎

  17. Alexandrov T, et al., Spatial Metabolomics and Imaging Mass Spectrometry in the Age of Artificial Intelligence, Annual Review of Biomedical Data Science, 13 April 2020.
    This review surveys spatial metabolomics and imaging mass spectrometry, which map the spatial distribution of metabolites across tissue sections while preserving anatomical context. It discusses the computational challenges of analyzing high-dimensional imaging datasets and the role of artificial intelligence in extracting biological insight from spatial metabolomic data. ↩︎