Python client for the OmniPath molecular biology prior-knowledge web API.
- Export entities, interactions, and associations (complexes, pathways, reactions) as DataFrames
- Ontology term lookup, search, and hierarchy trees
- Multi-backend output: polars (default), pandas, or pyarrow
- Optional graph conversion to annnet objects
- Query validation against the API schema
- Caching via download-manager
pip install omnipath-clientWith polars (recommended default backend):
pip install omnipath-client polarsimport omnipath_client as op
# All interactions as a polars DataFrame
df = op.interactions()
# Directed interactions only
df = op.interactions(direction='directed')
# Human entities
df = op.entities(taxonomy_ids=['9606'])
# Interactions as an annnet graph
g = op.interactions(as_graph=True)
# Ontology term lookup
result = op.ontology_terms(['GO:0006915', 'MI:0326'])
# Choose a different backend
df = op.entities(backend='pandas')For more examples, see the quickstart guide.
Full documentation: saezlab.github.io/omnipath-client
The data served by OmniPath is combined from many original resources, each with their own license terms. The OmniPath client software is BSD-3-Clause, but the data is subject to the licenses of the original sources. Some resources restrict commercial use. Tools and documentation for managing license-based access control will be provided in a future release.
If you use OmniPath in your research, please cite:
Türei D, Schaul J, Palacio-Escat N, Bohár B, Bai Y, Ceccarelli F, Çevrim E, Daley M, Darcan M, Dimitrov D, Doğan T, Domingo-Fernández D, Dugourd A, Gábor A, Gul L, Hall BA, Hoyt CT, Ivanova O, Klein M, Lawrence T, Mañanes D, Módos D, Müller-Dott S, Ölbei M, Schmidt C, Şen B, Theis FJ, Ünlü A, Ulusoy E, Valdeolivas A, Korcsmáros T, Saez-Rodriguez J. (2026) OmniPath: integrated knowledgebase for multi-omics analysis. Nucleic Acids Research 54(D1):D652-D660. doi:10.1093/nar/gkaf1126
The client software is licensed under BSD-3-Clause. See Data licensing above for information about the data.