Metabolomics tools from the Saez Lab and Korcsmaros Lab.
OmniPath Metabo brings together chemical structures, metabolite–protein interactions, annotations, cheminformatics and chemical ontologies into a single, harmonized prior-knowledge resource for metabolomics. It powers pre-built prior-knowledge networks for multi-omics causal reasoning (the COSMOS PKN), and is accessible through the web app, the HTTP web service, and the Python (omnipath-client) and R (OmnipathR, MetaProViz) packages.
Browse the OmniPath database in an interactive web app at dev.omnipathdb.org.
Access OmniPath data through our new web service at dev.omnipathdb.org/api/docs/.
Translate between dozens of chemical and gene product identifiers, and handle taxonomy and orthology seamlessly with OmniPath Utils.
Access the OmniPath web service directly from Python with omnipath-client. Available on PyPI and GitHub.
Direct, easy access to several metabolite databases from R with OmnipathR. Available on Bioconductor and GitHub.
Metabolomics data processing, differential and functional analysis and visualization in the R/Bioconductor package MetaProViz. Available on Bioconductor and GitHub.
These are the recommended entry points. For the full family of OmniPath and infrastructure packages, see More packages below.
OmniPath Metabo integrates diverse interactions, annotations, structures, cheminformatics, chemical ontologies and more. For the high-resolution PDF click here.
Workflow and design. OmniPath directly accesses original resources and harmonizes them to a common database schema. OmniPath data is available through 3 web APIs: the main OmniPath service, OmniPath Utils and OmniPath Metabo, all accessible through omnipath-client. For the high-resolution PDF click here.
The MetaProViz R/Bioconductor package is a toolkit for processing, differential and functional analysis and visualization of metabolomics data. It provides access to prior knowledge, however is not connected to OmniPath Metabo yet — but it will be soon. We will also deploy similar functionalities in Python. Read more in our preprint.
J Schaul, Y Bai, D Bottazzi, J Franken, T Lawrence, N Palacio-Escat, E Carreño,
M Daley, L Gul, A Sahin, D Mañanes, B Bohar, A Dugourd, T Korcsmaros, D Turei,
C Schmidt and J Saez-Rodriguez (2026) OmniPath Metabo: chemical structures,
interactions and mechanisms to study the metabolome.
bioRxiv
PDF |
DOI: 10.64898/2026.06.18.733117
C Schmidt, D Turei, D Prymidis, M Daley, C Frezza and J Saez-Rodriguez (2025)
Integrated metabolomics data analysis to generate mechanistic hypotheses with
MetaProViz.
bioRxiv
; accepted in Molecular Systems Biology
PDF |
DOI: 10.1101/2025.08.18.670781
D Türei, J Schaul, N Palacio-Escat, B Bohár, Y Bai, F Ceccarelli, E Çevrim, M Daley,
M Darcan, D Dimitrov, T Doğan, D Domingo-Fernández, A Dugourd, A Gábor, L Gul, BA Hall,
CT Hoyt, O Ivanova, M Klein, T Lawrence, D Mañanes, D Módos, S Müller-Dott, M Ölbei,
C Schmidt, B Şen, FJ Theis, A Ünlü, E Ulusoy, A Valdeolivas, T Korcsmáros and
J Saez-Rodriguez (2026) OmniPath: integrated knowledgebase for multi-omics analysis.
Nucleic Acids Res
54(D1): D652-D660
PMID: 41251164 |
PDF |
DOI: 10.1093/nar/gkaf1126
D Turei, A Valdeolivas, L Gul, N Palacio-Escat, M Klein, O Ivanova, M Olbei, A Gabor,
F Theis, D Modos, T Korcsmaros and J Saez-Rodriguez (2021) Integrated intra- and
intercellular signaling knowledge for multicellular omics analysis.
Mol Syst Biol
17: e9923
PMID: 33749993 |
PDF
D Turei, T Korcsmaros and J Saez-Rodriguez (2016) OmniPath: guidelines and gateway
for literature-curated signaling pathway resources.
Nature Methods 13 (12)
PMID: 27898060 |
Full text |
Supplementary
Not yet — please treat this as a preview. The new OmniPath database and its Metabo extension are still under active development. The contents are updated frequently and currently lack completeness and thorough quality control. Versioning and archiving are not implemented yet, and the hosting is not reliable. For these reasons, please use it for exploratory purposes only — it is not suitable for reproducible analysis. We expect it to become ready for that by autumn 2026. We welcome feedback and are happy to include new design ideas and resources — let us know!
The COSMOS prior-knowledge network (PKN) is a pre-built network of metabolite–protein interactions for multi-omics causal reasoning with cosmosR (preprint). An easy-to-use and much more powerful Python implementation is coming soon. The PKN combines transporters, receptors, allosteric regulators and enzyme–metabolite interactions, and can be filtered by category, resource and organism.
Which identifiers do you use?Metabolites are identified primarily by InChIKey, and also mapped to ChEBI. Gene products (including proteins) are identified primarily by NCBI Entrez Gene ID and UniProt. Records from different resources can therefore be combined directly. For translation between dozens of further chemical and gene product identifiers use OmniPath Utils.
Which organisms are supported?
The main OmniPath database and OmniPath Utils cover a wide range of organisms. The
specialized networks in the OmniPath Metabo web service are currently limited to human
and mouse. Every query takes an organism parameter (NCBI Taxonomy ID,
e.g. 9606 for human).
Through this HTTP web service with
interactive documentation, from Python with
omnipath-client, from R with
OmnipathR, or via the
omnipath-metabo Python package.
Our software is fully open source: the server-side components are licensed under GPLv3, while all client-side software is under the permissive, business-friendly BSD-3-Clause license, so it can be used freely in commercial projects. The OmniPath data itself carries no license of its own; instead, the database records the licenses of every integrated resource, and will soon support filtering its contents for for-profit (commercial) use. Please consult and cite the original resources when you use the data.
OmniPath Metabo is built and accessed through a family of open-source packages. The client packages are the recommended entry points; the others build the databases and web services or provide shared infrastructure.
| Package | Description | Docs | Code | Package |
|---|---|---|---|---|
| Client packages | ||||
| omnipath-client | Python client for all OmniPath web services (main, Utils and Metabo). | docs | GitHub | PyPI |
| OmnipathR | R/Bioconductor client for OmniPath: interactions, annotations and metabolite databases. | docs | GitHub | Bioconductor |
| MetaProViz | R/Bioconductor toolkit for metabolomics data processing, differential and functional analysis, and visualization. | docs | GitHub | Bioconductor |
| OmniPath software | ||||
| omnipath-resources (pypath) | Access and parsing of the original resources — the data-acquisition layer behind OmniPath. | docs | GitHub | PyPI |
| omnipath-build | Builds the main OmniPath database from the integrated resources, on top of pypath. | — | GitHub | — |
| omnipath-present | The OmniPath web application: interactive database browser and HTTP API service. | — | GitHub | — |
| omnipath-utils | Identifier translation, taxonomy, orthology and reference lists for molecular biology. | docs | GitHub | PyPI |
| omnipath-metabo | OmniPath Metabo build extension and metabolomics web API (COSMOS PKN, cheminformatics, chemical ontologies). | docs | GitHub | — |
| ontosparse (ontograph) | Sparse ontology engine running inside the web service: loads, represents and efficiently queries biological ontologies as graph structures. | docs | GitHub | PyPI |
| Infrastructure | ||||
| dlmachine | Download manager: robust, resumable downloads of remote resources. | — | GitHub | PyPI |
| cachedir | Cache manager: organizes on-disk caches of downloaded and processed data. | docs | GitHub | PyPI |
| pkg-infra | Shared session, configuration and logging infrastructure for Saez Lab packages. | docs | GitHub | PyPI |
| Reproducibility | ||||
| metabo-usecases | Integrated pipeline reproducing the figures, tables and case studies of the OmniPath Metabo paper. | — | GitHub | — |
Jonathan Schaul, Yunfan Bai, Daniele Bottazzi, Jannik Franken, Toby Lawrence, Nicolàs Palacio-Escat, Edwin Carreño, Macabe Daley, Lejla Gul, Ardic Sahin, Diego Mañanes, Balazs Bohar, Aurelien Dugourd, Tamas Korcsmaros, Denes Turei, Christina Schmidt and Julio Saez-Rodriguez.
