Metabolomics across scales: from single cells to population studies
Summary
This Nature review charts the rapid expansion of metabolomics methods and applications across biological scales — from single-cell and spatial approaches to high-throughput population studies. It highlights recent technical advances in mass spectrometry, imaging and multiplexing that enable detection of small-molecule metabolites at cellular and subcellular resolution, plus large-cohort efforts that show metabolite profiles can predict disease and ageing. The article also discusses integration with other omics, isotope tracing for flux analysis, computational and annotation challenges, and the clinical and translational potential of metabolomics.
Key Points
- Single-cell and spatial metabolomics technologies (DESI-MSI, MALDI, SpaceM, probe-based MS) are rapidly maturing, revealing metabolic heterogeneity within tissues and cell types.
- High-throughput single-cell methods (e.g. HT SpaceM, isobaric multiplexing) are increasing scale and reproducibility for small-molecule detection.
- Spatial multimodal methods now allow concurrent detection of transcriptomes and metabolomes from the same tissue section, improving cell-state annotation.
- Isotope tracing and flux analysis remain essential for linking metabolite levels to pathway activity and mechanism rather than mere association.
- Population-scale metabolomics (large biobanks, UK Biobank, Nightingale studies) demonstrate strong predictive power for multi-disease outcomes and biological ageing — sometimes outperforming polygenic scores.
- Integration with genomics, proteomics and immune phenotypes is uncovering causal roles of metabolites in disease and informing drug-mode-of-action maps.
- Machine learning and foundation-model approaches (METASPACE-ML, DreaMS, other ML tools) are improving annotation and interpretation of complex mass spectra and imaging data.
- Major challenges remain: confident metabolite identification, absolute quantitation, batch normalisation, throughput trade-offs and standardised computational workflows for atlas-level integration.
Why should I read this?
Quick answer: if you care about how metabolism actually works — not just at the level of whole blood but right down inside individual cells and across whole populations — this review pulls it together. It saves you time by flagging the key methods, big cohort wins and the sticky problems (annotation, quantitation, integration) so you know where the field is actually moving — and where to focus your next experiment or grant bid.
Author style
Punchy — the review compresses a huge, fast-moving literature into a single narrative. If you work in metabolomics, single-cell biology, clinical translation or computational biology this piece amplifies why reading the details matters: the technical innovations and large-cohort results together push metabolomics from biomarker hunting toward mechanism, prediction and actionable medicine.
Context and Relevance
Why it matters: advances in single-cell and spatial metabolomics are changing how we understand cellular heterogeneity, tissue microenvironments and cell–cell metabolic interactions (important in cancer, immunity and regeneration). At the same time, population-scale metabolomics demonstrates clinical and epidemiological utility — predicting disease risk, biological age and treatment responses. The convergence of high-resolution measurement, isotope-based flux analysis, multimodal integration and ML-driven annotation is positioning metabolomics as a central pillar of precision medicine and systems biology. However, adoption requires standardised protocols, improved identification pipelines and robust normalisation to translate discoveries into the clinic.