Skip to content

Metabolomics

Untargeted metabolomics at cohort scale.

Thousands of samples, millions of reference spectra, one alignment. From raw LC-MS, GC-MS or ion mobility files to an annotated feature table you can actually interpret.

Discuss your workflow All applications

The bottleneck

Where metabolomics stops scaling

A pilot of thirty samples works in almost any tool. A cohort of three thousand does not. Alignment drifts, annotation slows to the point where nobody checks it properly, and processing quietly becomes the rate-limiting step between acquisition and biology.

  • Most software was never designed for cohort-scale batches
  • Annotation that is either fast or trustworthy, rarely both
  • Weeks between the last injection and the first result

The workflow

From raw files to something you can publish.

Every step ships with default settings. You may customise each one, but you don’t have to.

  • Import

    Native readers for every major vendor, DDA and DIA, with no conversion step.

  • Detect and resolve

    Mass detection, chromatogram building and deconvolution, each with default settings you can customise.

  • Ion mobility

    If you acquire it, the mobility dimension is carried through as data rather than reduced to a column.

  • Align the cohort

    One feature table across every sample, matched on m/z, retention time, mobility and MS2 similarity.

  • Ion identity networking

    Adducts, in-source fragments and isotopologues collapsed onto single molecular identities.

  • Annotate

    Reference spectra, compound databases and prediction tools where the library ends.

Capabilities

What you get out of the box.

Annotated mass spectrum with a matched molecular structure and spectral library lookup

LC-MS2, DDA and DIA

Analyse your LC-MS2 DDA and DIA Metabolomics data.

Spectral matching at scale

Match millions of reference spectra against 1000s of samples within seconds.

Advanced spectra merging

Maximum control with advanced spectra merging options

GC-MS profiling

Analyse your GC/CI and GC/EI profiling data. Fast GC-EI/MS spectral deconvolution and feature alignment. Apply molecular networking to your GC-MS data. Interactive spectral deconvolution result viewer

In the software

The compound dashboard.

An annotation is only worth the evidence behind it. Here erucamide, a plasticware additive, and the kind of thing you want caught rather than reported, is checked against the blank and scored: two adducts present, retention time stable to 0.005 minutes, thirty MS2 scans behind every assessment of the compound, no in-source fragments muddying the assignment. All 288 compound rows in the list carry the same panel, which is what makes triage across a cohort possible at all.

Metabolomics
The mzmine compound dashboard: a retention time against m/z map of the aligned feature list, chromatographic peak shapes for the [M+H]+ and [M+Na]+ adducts, the MS1 and merged MS2 spectra, a compound quality panel listing the checks the erucamide annotation passed, and the feature table below with 288 compound rows, blank-against-sample height plots and drawn structures

In production

Labs already running this.

“As part of the R&D scent team at IFF, working with complex and diverse natural product extracts, efficient data processing, deconvolution, annotation, chemometrics, and interactive visualization are crucial for high-throughput understanding. mzmine Pro has provided us with not only a powerful and efficient GC-MS workflow to investigate volatile compounds but also a highly advanced LC-MS workflow, perfectly aligned with our aspirations. These tools are essential for enhancing our work, fostering innovation, and deepening our understanding of complex natural matrices.”
Dr. Melissa Nothias-Esposito
Dr. Melissa Nothias-Esposito
Senior Scientist · LMR by IFF, Grasse, France
“My lab develops and applies bioanalytical tools to determine the role of small molecules in complex ecosystems. Mass Spectrometry is our tool of choice to profile large environmental studies, often exceeding 1000 samples per study. With mzmine, we can analyze the vast amount of data in less than an hour, giving us time to focus on the actual biology. Our current processing record of ultra-complex dissolved organic matter samples is more than 8000 samples in 45 min.”
Dr. Daniel Petras
Dr. Daniel Petras
Assistant Professor of Biochemistry · University of California Riverside, USA

The science

The methods behind this workflow, peer-reviewed.

2023
Integrative analysis of multimodal mass spectrometry data in MZmine 3 (opens in a new tab)
Schmid et al. · Nature Biotechnology 41, 447–449
The methods paper for the platform itself. Cite this one if you cite only one.
2026
Self-supervised learning of molecular representations from millions of tandem mass spectra using DreaMS (opens in a new tab)
Bushuiev et al. · Nature Biotechnology 44, 630–640
A foundation model for MS2 spectra, trained on millions of unannotated spectra.
2025
MSnLib: efficient generation of open multi-stage fragmentation mass spectral libraries (opens in a new tab)
Brungs, Schmid, Heuckeroth et al. · Nature Methods 22, 2028–2031
How the in-house libraries are built, and why an MSn library beats a flat MS2 one.
2021
Ion identity molecular networking for mass spectrometry-based metabolomics in the GNPS environment (opens in a new tab)
Schmid et al. · Nature Communications 12, 3832
Resolves adducts and in-source fragments into single molecular identities.
2020
Feature-based molecular networking in the GNPS analysis environment (opens in a new tab)
Nothias, Petras, Schmid et al. · Nature Methods 17, 905–908
The networking method the drug discovery and dereplication workflows rest on.

Resources

Go deeper on Metabolomics

We have written this up in more detail. There are one paper and two posters. One short form and all of it unlocks.

Request the material

Get started

See it on your own data.

Send us a few representative files. We will build the metabolomics workflow and show you the result before you commit to anything.

Get in touch Compare plans