is an Aggregator.
It is part of the Translator collection.
Babel is the NCATS Biomedical Data Translator pipeline, built at RENCI, that groups equivalent identifiers from many biomedical vocabularies into cliques and assigns each clique a preferred identifier, label and Biolink type. Release 2026jul22 holds about 605.9 million CURIEs in 388.5 million cliques, built against Biolink Model 4.4.3. Its compendia, synonym and conflation files back the Translator Node Normalizer and Name Resolver services.
No external IDs available
| ID | Name | URL | Category | Format | Description |
|---|---|---|---|---|---|
| babel.compendia | Babel Compendia | compendia | Product | json | Babel compendia, one file per Biolink... |
| babel.synonyms | Babel Synonyms | synonyms | Product | json | Synonym files for each Babel type, on... |
| babel.conflation | Babel Conflation Files | conflation | Mapping | json | Conflation files that merge cliques a... |
| babel.kgx | Babel KGX Files | kgx | Graph | kgx-jsonl | Babel compendia in KGX JSON Lines for... |
| babel.duckdb | Babel DuckDB and Parquet Exports | duckdb | Product | parquet | DuckDB databases and Apache Parquet e... |
| babel.code | Babel Source Code | Babel | Process | python | Source code for the Babel pipeline (S... |
| babel.docs | Babel Data Formats Documentation | DataFormats.md (5.8 KB) | Documentation | http | Babel documentation on the download l... |
| ID | Name | URL | Category | Format | Relation | Description |
|---|---|---|---|---|---|---|
| name-resolver.api | Name Resolver API (RENCI) | docs | Programming Interface | http | was derived from | RENCI-hosted Name Resolver REST API (... |
| name-resolver.translator-api | Name Resolver API (Translator production) | docs | Programming Interface | http | was derived from | NCATS Translator production deploymen... |
| node-normalizer.api | Node Normalization API (RENCI) | docs | Programming Interface | http | was derived from | Node Normalization REST API hosted by... |
| node-normalizer.api.translator | Node Normalization API (Translator) | docs | Programming Interface | http | was derived from | Node Normalization REST API deployed ... |
Babel builds cliques of equivalent identifiers across biomedical vocabularies for the NCATS Biomedical Data Translator. It reads identifiers and cross-references from sources such as UniProtKB, PubChem, NCBI Gene, Ensembl, UMLS, ChEMBL, MeSH, ChEBI, the OBO ontologies and DrugBank, merges them into cliques by Biolink type, and picks a preferred identifier and label for each clique. RRID: SCR_028595.
Downloads live at https://stars.renci.org/var/babel_outputs/, with one directory per release. The latest/ directory pointed to release 2026jul22 when checked on 2026-10-04 (its VERSION.txt says so). That release was built with Babel about v1.18.1 against Biolink Model 4.4.3 and holds 605,864,191 CURIEs in 388,490,111 cliques. The release notes are in the repository under releases/2026jul22/. UniProtKB, PubChem compounds, InChIKeys and NCBI Gene make up most identifiers.
Each release contains compendia/, synonyms/, conflation/, kgx/ and duckdb/ directories, plus reports, benchmarks and logs. The compendia and conflation files use a custom JSON Lines format described in docs/DataFormats.md, even though they carry a .txt extension.
Babel outputs are loaded into the Translator Node Normalizer (NodeNorm), which maps identifiers to preferred CURIEs, and the Name Resolver (NameRes), which looks up concepts by name.
The code is under the MIT license. The release outputs carry no separate data license, and individual sources keep their own terms.
Created: October 04, 2026 | Last modified: October 04, 2026