omnipath-client
Python client for the OmniPath molecular biology prior-knowledge web API.
omnipath-client provides validated, cache-aware access to the OmniPath API, delivering results as polars, pandas, or pyarrow DataFrames. Network data can optionally be returned as annnet graph objects.
Features
- Export endpoints for entities, interactions, and associations (complexes, pathways, reactions)
- Ontology endpoints for term lookup, search, and hierarchy trees
- Multi-backend DataFrame output: polars (default), pandas, pyarrow
- Query validation against the API schema (endpoint names, parameter names, enum values)
- Caching of downloaded data via download-manager
- Graph conversion to annnet objects for network data
Quick example
import omnipath_client as op
# Get all interactions as a polars DataFrame
df = op.interactions()
# Filter by direction
df = op.interactions(direction='directed')
# Get entities for a specific organism
df = op.entities(taxonomy_ids=['9606'])
# Return interactions as an annnet graph
g = op.interactions(as_graph=True)
Getting started
- Installation -- how to install the package
- Quickstart -- basic usage examples
- API Reference -- full API documentation