chembl-mcp-server

Drug-discovery data over ChEMBL (EBI) — the curated link between compounds, protein targets, and measured bioactivity (IC50/Ki/EC50), plus drug mechanisms and indications. Canonical chains: (1) a UniProt accession from the uniprot/protein server → chembl_search_targets → chembl_get_bioactivities for the most potent leads on a target, or with molecule_chembl_id supplied alongside target_chembl_id to narrow to how potently one compound hits that one target; (2) chembl_search_molecules → chembl_get_drug_info for a drug's mechanism and indications; (3) a molecule's standard_inchi_key → the pubchem server for richer chemistry, an approved drug (max_phase 4) → the openfda server for the FDA label and adverse events. Ranking trap: pchembl_value is comparable only within one standard_type — set the standard_type filter (mixing IC50 and Ki is a scientific error). Coverage trap: many measurements have no derivable pchembl_value and are absent from the default potency-ranked view