OmniPath Metabolomics provides pre-built prior-knowledge networks for multi-omics causal reasoning (COSMOS). The COSMOS PKN combines metabolite–protein interactions from TCDB, SLC, BRENDA, MRCLinksDB, STITCH, genome-scale metabolic models (Human-GEM, Recon3D), and more. Available as a Python library and as this HTTP API with interactive documentation.

• COSMOS PKN — pre-built prior-knowledge network combining transporters, receptors, allosteric regulators, and enzyme–metabolite interactions
• ID translation — all metabolites unified to ChEBI, all proteins to primary SwissProt ACs
• Multiple resources — TCDB, SLC, BRENDA, MRCLinksDB, STITCH, Human-GEM, Recon3D
• Multi-organism — human, mouse, and additional organisms (coming soon)

COSMOS PKN API

Get the full human transporter PKN:

Get transporters + receptors:

Filter by resource:

More endpoints:

• /cosmos/categories — available PKN categories
• /cosmos/organisms — organisms with pre-built PKNs
• /cosmos/resources — resources per category
• /cosmos/status — cache status and build info

Python usage

pip install omnipath-metabo

from omnipath_metabo.datasets.cosmos import build_transporters
bundle = build_transporters()
# 51,067 edges, 3,045 metabolites, 2,317 proteins