Written and maintained by CASRAI Editorial Board
Last updated
MetaboAnalyst is a web-based platform for metabolomics data processing, statistical analysis, functional interpretation, and multi-omics integration. Researchers handling liquid chromatography–mass spectrometry (LC-MS), gas chromatography–mass spectrometry (GC-MS), or nuclear magnetic resonance (NMR) datasets face steep computational hurdles, moving from raw spectra to high-dimensional feature matrices. MetaboAnalyst bridges instrumentation outputs and biological insights through structured graphical workflows powered by reproducible backend R code.
What It Is
MetaboAnalyst functions as a centralized bioinformatics suite for processing targeted and untargeted metabolomics data. The platform provides eighteen analytical modules spanning raw spectral deconvolution, quality control filtering, missing value imputation, multivariate pattern recognition, pathway enrichment, and multi-omics integration.
What Is Metabolomic Technology?
Metabolomic technology encompasses analytical chemistry platforms and software used to detect, identify, and quantify small molecules (metabolites with molecular weights typically under 1,500 Daltons) within a biological system. The primary instrumentation platforms include:
- Liquid Chromatography–Mass Spectrometry (LC-MS): Combines chromatographic separation with high-resolution mass detection, providing broad metabolome coverage. In untargeted experiments, LC-MS generates complex profiles requiring retention time alignment and quality control samples, as detailed in our guide to mass spectrometry platform choice and QC design.
- Gas Chromatography–Mass Spectrometry (GC-MS): Ideal for volatile and derivatized metabolites such as organic acids and amino acids, utilizing reproducible electron ionization libraries.
- Nuclear Magnetic Resonance (NMR) Spectroscopy: Provides non-destructive quantitative measurements without chromatographic separation, though with lower sensitivity than mass spectrometry.
What Is the Difference Between Metabonomics and Metabolomics?
While often used interchangeably in contemporary literature, the two terms possess distinct historical origins:
- Metabonomics: Defined in 1999 by Jeremy Nicholson and colleagues, metabonomics measures dynamic multiparametric metabolic responses of living systems to pathophysiological stimuli or genetic modification, historically emphasizing biofluid NMR profiling.
- Metabolomics: Defined around the same era by Stephen Oliver, Oliver Fiehn, and peers, metabolomics aims for the comprehensive identification and quantification of all low-molecular-weight metabolites in a biological specimen across analytical platforms.
In modern practice, metabolomics serves as the standard overarching term covering both targeted quantification and global untargeted discovery pipelines.
Who Runs It and What It Covers
MetaboAnalyst is maintained by the Xia Lab at McGill University (led by Dr. Jianguo Xia), with contributions from the University of Alberta (Wishart Lab) and international collaborators. The public platform is hosted at the McGill University Data Center. Computations are powered by the R package MetaboAnalystR (version 4.0) and optiLCMS. The software is licensed under the GNU General Public License (GPL v2 or later), freely available for academic and commercial use.
The platform organizes 18 modules across 5 primary input pathways:
- LC-MS Spectra: Centroided raw spectra processing (mzML, mzXML, NetCDF) using the asari algorithm for high-resolution LC-MS.
- MS Peaks: Tandem mass spectrometry (MS2) fragment annotation against reference spectral libraries, and peak-to-pathway functional analysis via mummichog.
- Generic Formats: Statistical analysis of concentration and peak intensity tables (single-factor and multi-factor metadata designs), biomarker evaluation, and dose-response calculations.
- Annotated Features: Over-representation enrichment and pathway topology analysis across approximately 130 species, mapped to the KEGG pathway database and the Reactome pathway knowledgebase.
- Genomic and Phenotypic Links: Two-sample Mendelian randomization connecting metabolite quantitative trait loci with phenotypic traits for causal inference.
How to Search or Use It
How to Analyze Untargeted Metabolomics Data?
Untargeted metabolomics workflows convert raw spectral scans into biologically interpretable compounds. Key phases include open format conversion (to mzML), peak deconvolution, quality control (QC) filtering, normalization, multivariate modeling, and MS2 spectral library matching. Specialized tools such as GNPS molecular networking are frequently combined with MetaboAnalyst to group unknown spectral families.
Within MetaboAnalyst, researchers execute this pipeline through seven sequential interface steps:
Step 1: Module Selection
On the Module Overview page, choose your analytical track: select Statistical Analysis [one factor] for standard group comparisons, Statistical Analysis [metadata table] for multi-factor clinical studies with covariates, or Spectra Processing [LC-MS1 w/wo MS2] for raw data extraction.
Step 2: Data Upload
In the Please upload your data interface, upload a formatted CSV or TXT file (concentrations, spectral bins, or peak intensities). Include pooled quality control samples labelled QC_### (class QC) and blanks labelled BLANK_### (class BLANK), or directly import public datasets via a Metabolomics Workbench Study ID.
Step 3: Data Integrity Check
The system evaluates uploaded matrices for non-numeric entries, negative values, constant features, and missing values, displaying a summary verification panel.
Step 4: Missing Value Imputation and Data Filtering
In Missing Value Imputation, replace missing cells with half the minimum detected positive value, or select algorithms including KNN, PPCA, or BPCA. The Data Filtering step removes uninformative variables using interquartile range (IQR) or relative standard deviation (RSD) thresholds evaluated across pooled QC samples.
Step 5: Data Normalization
The Normalization view applies row-wise sample normalization (by median, sum, or reference feature), column-wise transformation (log10 or cube root), and column-wise scaling. Researchers typically select Pareto scaling to reduce dynamic range while preserving biological contrast, or Auto scaling (unit variance).
Step 6: Statistical Analysis
Explore results using univariate tools (fold change, Student t-tests, ANOVA, volcano plots with FDR-adjusted p-values) and multivariate methods (PCA score plots with confidence ellipses, PLS-DA with VIP scores, and hierarchical clustering heatmaps).
Step 7: Automated Report Generation
Click Generate Report to compile a publication-ready PDF documenting analysis parameters, equations, and figures, alongside a downloadable R script (MetaboAnalystR.R) enabling local script-based computational verification.
Where to Deposit Metabolomics Data?
When choosing an open data repository, metabolomics researchers primarily use two archives: MetaboLights (EMBL-EBI, assigning MTBLS##### accessions using ISA-Tab/mzTab-M metadata) and the Metabolomics Workbench (NIH NMDR, assigning ST###### study IDs). Accession codes should be documented in the manuscript’s data availability statement.
Strengths and Limits
Strengths
- Broad Analytical Scope: Integrates raw spectra processing, multivariate statistics, pathway enrichment, and multi-omics within a unified browser interface.
- GUI Usability with Script Reproducibility: Offers code-free web interaction while exporting underlying R command scripts for independent execution.
- Standardized Ingestion: Natively accepts open community formats including mzML, mzTab 2.0-M, and Metabolomics Workbench study IDs.
Limits
- 72-Hour Server Retention: User directories at the McGill Data Center are temporary and automatically deleted by scheduled routines within 72 hours. Output files must be downloaded promptly.
- Discontinued Local Web Server: Local web server installations have been discontinued. Workflows requiring local execution must use the
MetaboAnalystRR package. - Bandwidth Constraints: Uploading large raw LC-MS batches through web forms can experience transmission timeouts during peak server demand.
Alternatives
Researchers select metabolomics software according to computational architecture and experimental design:
| Platform | Primary Focus | Deployment Model | User Interface | Key Strengths |
|---|---|---|---|---|
| MetaboAnalyst 6.0 | Statistical analysis, pathway enrichment, and multi-omics | Cloud web server and R package | Web browser GUI and R CLI | 18 modules, automated PDF reports, causal Mendelian randomization. |
| XCMS Online | Raw LC-MS/GC-MS peak extraction and pairwise matching | Cloud web server | Web browser GUI | Cloud-based feature extraction and retention time alignment; comparative statistics. |
| MZmine 3 | Raw MS1/MS2 spectral deconvolution and feature extraction | Local desktop application (Java) | Desktop GUI | Modular processing for untargeted LC-MS and ion mobility data; interactive peak visualization. |
| MS-DIAL | Untargeted metabolomics and lipidomics deconvolution | Local desktop application (Windows .NET) | Desktop GUI | Extensive in silico and experimental MS/MS library matching; specialized lipidomics tools. |
| GNPS | Tandem mass spectrometry (MS/MS) molecular networking | Cloud web platform | Web browser GUI | Crowd-sourced MS/MS spectral matching, substructure clustering, and natural product discovery. |
Frequently Asked Questions
What is metabolomic technology?
Metabolomic technology refers to analytical chemistry platforms—primarily LC-MS, GC-MS, and NMR spectroscopy—used to identify and quantify small-molecule metabolites under 1,500 Daltons in biological specimens.
What is the difference between metabonomics and metabolomics?
Metabonomics was defined in 1999 as the measurement of dynamic metabolic responses to pathophysiological stimuli, historically focusing on biofluid NMR. Metabolomics aims for the comprehensive analysis of all metabolites within a biological system. Both terms are frequently used interchangeably today.
How to analyze untargeted metabolomics data?
Untargeted metabolomics requires converting raw spectra to open centroided formats, performing peak deconvolution, filtering unstable features with pooled QC samples, applying normalization and scaling, conducting multivariate modeling (PCA, PLS-DA), and annotating discriminating features against MS/MS libraries.
Where to deposit metabolomics data?
Metabolomics datasets should be deposited in public archives such as MetaboLights (EMBL-EBI, MTBLS accessions) or Metabolomics Workbench (NIH NMDR, ST study IDs). Tandem mass spectrometry data can also be submitted to GNPS and MassIVE.
How should MetaboAnalyst 6.0 be cited?
Cite the primary reference: Pang, Z., Lu, Y., Zhou, G., Hui, F., Xu, L., Viau, C., Spigelman, A. F., MacDonald, P. E., Wishart, D. S., Li, S., and Xia, J. (2024). MetaboAnalyst 6.0: towards a unified platform for metabolomics data processing, analysis and interpretation. Nucleic Acids Research, 52(W1), W398–W406 (doi: 10.1093/nar/gkae253).
Sources
- MetaboAnalyst Platform Overview and Documentation. Xia Lab, McGill University. https://www.metaboanalyst.ca (Read November 7, 2026).
- MetaboAnalyst Data Formats and Specifications. https://www.metaboanalyst.ca/docs/Format.xhtml (Read November 7, 2026).
- MetaboAnalyst Server and Data Policy. https://www.metaboanalyst.ca/docs/DataPolicy.xhtml (Read November 7, 2026).
- Pang, Z., Lu, Y., Zhou, G., et al. (2024). MetaboAnalyst 6.0: towards a unified platform for metabolomics data processing, analysis and interpretation. Nucleic Acids Research, 52(W1):W398–W406. doi: 10.1093/nar/gkae253 (Read November 7, 2026).
- MetaboLights Repository. European Bioinformatics Institute (EMBL-EBI). https://www.ebi.ac.uk/metabolights/ (Read November 7, 2026).
- Metabolomics Workbench. National Metabolomics Data Repository (NMDR). https://www.metabolomicsworkbench.org/ (Read November 7, 2026).
