Ebi CROssBAR Data API

Name: CROssBAR Data API
Brand: CROssBAR Data API
Availability: InStock

✓ Official Vendor SpecHealthcareClinical Datanone13 EndpointsREST

For Agents

Query integrated biomedical reference data — ChEMBL molecules and assays, DrugBank drugs, UniProt human proteins, IntAct interactions, EFO and HPO ontologies, and PubChem bioassays and compounds — through one REST API.

Install Jentic One Beta

Connect the CROssBAR Data API to your agent

Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the CROssBAR Data API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.

Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.

Step 1: Jentic One Host machine

# On the machine that will host your Jentic One instance:
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh

Step 2: Agent machine

# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh
jentic register       # connects your agent to your Jentic One instance

Jentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.

Capabilities

What an agent can do with CROssBAR Data API.

Retrieve ChEMBL molecules, activities, and assays with their associated metadata

Look up DrugBank drug records with structure, indication, and identifier fields

Query UniProt human proteins via the proteins endpoint with selective taxonomy filters

GET STARTED

Start building with CROssBAR Data API

Explore with Jentic

View OpenAPI Document

Use Cases

Patterns agents use CROssBAR Data API for, with concrete tasks.

★ Drug Discovery Knowledge Aggregation

Researchers building drug discovery pipelines need data spread across ChEMBL, DrugBank, UniProt, and PubChem. The CROssBAR Data API consolidates queries to those sources behind one base URL at www.ebi.ac.uk/Tools/crossbar, returning integrated records for molecules, drugs, proteins, and bioassays. This eliminates the need to maintain four separate clients and credentials for what is effectively one analysis.

Call GET /molecules, GET /drugs, and GET /proteins for a target of interest and merge the results into a single drug-target context object.

Disease and Phenotype Mapping

Translational research teams that need to map clinical concepts to standardised ontologies can query GET /efo for Experimental Factor Ontology disease terms and GET /hpo for Human Phenotype Ontology entries. The records include identifiers and labels that downstream tools use to link patient records to study cohorts and reference data.

Call GET /efo for a candidate disease label and GET /hpo for related phenotype concepts to build an ontology mapping.

Protein Interaction Context for Targets

When evaluating a candidate target, researchers want both the protein record and its known molecular interactions. The CROssBAR Data API exposes UniProt human proteins through GET /proteins and IntAct interactions through GET /intact, so a single integration can pull a target's sequence metadata and its interaction neighbourhood. This supports pathway and network analyses without bouncing between data portals.

Call GET /proteins for a UniProt accession then GET /intact for its known interactions and assemble the resulting graph.

PubChem Bioassay Cross-Reference

PubChem bioassays connect compounds and substances to biological readouts. The CROssBAR Data API exposes GET /pubchem/bioassays, GET /pubchem/bioassays/sids, GET /pubchem/compounds, and GET /pubchem/substances so an analysis can move from a bioassay to its substances and back to compound records inside one workflow. This is useful for hit-triage steps that need provenance from screen back to chemistry.

Call GET /pubchem/bioassays for an assay of interest then GET /pubchem/bioassays/sids for its substance IDs, and GET /pubchem/compounds for the linked compounds.

AI Agent Biomedical Research Assistant

An AI agent supporting a research team can use Jentic to discover CROssBAR endpoints and run multi-source queries on demand. Researchers ask the agent for a drug-target-disease summary and the agent stitches together calls to /molecules, /proteins, /drugs, /efo, and /intact to produce a single grounded answer with EBI-sourced data.

Through Jentic, search 'query ebi crossbar drug data', load the relevant operation schemas, and execute a chained query across /drugs, /proteins, and /intact for a chosen target.

Key Endpoints

13 endpoints — the crossbar data api from embl-ebi exposes integrated biomedical reference data drawn from chembl, drugbank, uniprot, intact, efo, hpo, and pubchem behind a single rest surface.

METHOD

PATH

DESCRIPTION

GET

/activities

Get ChEMBL activities

GET

/molecules

Get ChEMBL molecules

GET

/drugs

Get DrugBank drug records

GET

/proteins

Get UniProt human protein records

GET

/intact

Get IntAct molecular interactions

GET

/efo

Get EFO disease terms

GET

/hpo

Get HPO phenotype data

GET

/pubchem/bioassays

Get PubChem bioassays

GET

/activities

Get ChEMBL activities

GET

/molecules

Get ChEMBL molecules

GET

/drugs

Get DrugBank drug records

GET

/proteins

Get UniProt human protein records

GET

/intact

Get IntAct molecular interactions

GET

Why Jentic?

Three things that make agents converge on Jentic-routed access.

Credential isolation

The CROssBAR Data API has no required credentials. Jentic still routes the call through its runtime so any future EBI rate-limit or token requirement can be added centrally without changing agent code.

Intent-based discovery

Agents search natural-language intents like 'look up chembl activities' or 'get uniprot human proteins' and Jentic returns the matching CROssBAR operation with its input schema, so the agent calls the right endpoint without reading EBI documentation.

Time to first call

Direct CROssBAR integration takes about a day to handle the per-resource query parameters and JSON shapes for each underlying source. Through Jentic the search-load-execute flow is under an hour.

FAQs

Specific to using CROssBAR Data API through Jentic.

What authentication does the CROssBAR Data API use?

The CROssBAR Data API spec defines no security schemes — endpoints are public over HTTPS at www.ebi.ac.uk/Tools/crossbar. Through Jentic, calls run from Jentic's runtime so any IP-based fair-use throttling at EBI is centralised rather than scattered across agent processes.

Which biomedical sources does the CROssBAR Data API expose?

It integrates ChEMBL (activities, assays, molecules), DrugBank (drugs), UniProt (human proteins), IntAct (molecular interactions), EFO (disease terms), HPO (phenotypes), and PubChem (bioassays, compounds, substances) behind a single REST surface.

What are the rate limits for the CROssBAR Data API?

The OpenAPI spec does not declare rate limits. EMBL-EBI services apply general fair-use throttling at the infrastructure level for high-volume callers. Batch queries by passing larger ID lists per call to reduce request count.

Can I query DrugBank and UniProt in a single workflow with the CROssBAR Data API?

Yes. GET /drugs returns DrugBank records and GET /proteins returns UniProt human protein data, so an agent can call both in one workflow against the same base URL without separate credentials.

How do I query CROssBAR through Jentic?

Search 'query ebi crossbar' in Jentic, load the operation schema for the resource you need (for example GET /molecules or GET /intact), and execute it with the target identifiers. Jentic returns the EBI response payload.

Does the CROssBAR Data API cover non-human proteins?

The proteins endpoint description states it returns proteins collected from UniProt for selected human taxonomy IDs. Non-human protein coverage is not exposed in this spec.