The problem. Metabolites — the small-molecule end products of cellular activity — vary with age, sex, and body mass, but untangling those effects from a noisy untargeted LC-MS dataset is a full analysis problem: normalisation, quality control, multivariate modelling, and honest validation, all before you trust a single association.

The idea. Thévenot and colleagues profiled urine from a healthy cohort and worked through that pipeline end to end — signal drift correction, sample/feature filtering, and PLS/OPLS modelling to relate the metabolome to age, BMI and gender with cross-validated performance. The value is as much the methodology demonstrated as the specific associations found.

Why it matters. This rounds out the omics tour: after genomics (variants), transcriptomics (RNA-seq), and proteomics (MaxQuant/MaxLFQ), metabolomics is the fourth layer, and it has the same shape — messy instrument data, systematic drift to correct, multivariate structure to model. The sacurine dataset became a teaching standard (Workflow4Metabolomics, ropls), so it’s the on-ramp if I ever touch this modality.

Verdict. Solid and pedagogically valuable more than groundbreaking — its lasting role is as a reference dataset and a template for untargeted metabolomics QC. Read it for the pipeline discipline: drift correction and validation are where metabolomics analyses live or die.