Hi everybody,
I am pleased to announce that Babel 2026jul22 [1] is now available! If you are a member of the Translator project, I’ve shared the download links in our Slack channel; if not, I plan to create a publicly accessible version of this release that you can download in the next few weeks. Babel 2026jul22 is available for testing on NodeNorm Experimental (https://nodenormalization-exp.apps.renci.org/docs) and NameRes Experimental (https://name-resolution-exp.apps.renci.org/docs). Unless I hear from you by Wednesday, I plan to update NodeNorm Dev [2] and NameRes Dev [3] next Thursday, August 20, 2026 starting around 1pm ET, with 2-4 hours of downtime expected. The APIs have not changed significantly, but you may want to check if your software works against these new endpoints before I switch, and please let me know if I shouldn’t do the upgrade next week (e.g. if you have a significant presentation or need stability around that time).
Here is a list of the most significant changes in this release — there are a LOT of them since this is the first Babel release since 2026sep1 (the current NodeNorm on Translator CI):
- Lots of substantial changes<https://github.com/NCATSTranslator/Babel/tree/main/releases/2026jul22#areas-that-changed-substantially> to Proteins (big reduction from UniProtKB's removal of non-reference proteins<https://www.uniprot.org/help/refprot_only_changes>), chemicals (mostly reorganizations from small molecule from/to other chemical entities), MacromolecularComplex (add all complexes from ComplexPortal, not just S. cerevisiae)
- Added the Mammalian Phenotype Ontology<https://github.com/NCATSTranslator/Babel/pull/886> and the EMAPA Mouse Developmental Anatomy Ontology<https://github.com/NCATSTranslator/Babel>, as well as tooling to judge the effect of adding new sources to Babel
- DrugBank entries that are food or complex molecular mixtures have been retyped as food (but see one outstanding bug<https://github.com/NCATSTranslator/Babel/issues/962>)
- Improved preferred labels and fixed several bugs in gene names and synonyms, ChEMBL labels and others.
- ChEBI secondary identifiers and PubChem xrefs have been restored
- Simplified how we choose DrugChemical conflation clique leaders
- Metadata files<https://github.com/NCATSTranslator/Babel/tree/main/releases/2026jul22/metadata> now provide some information on the provenance of each of the output files -- all xrefs used should be included here, information on individual downloads coming later this year<https://github.com/NCATSTranslator/Babel/issues/1013>
- NodeNorm Redis loaders and NameRes loaders are now much faster, and NodeNorm Redis can support recent Solr versions.
- NameRes searches using only_taxa no longer filter out identifiers that are not taxon specific (example<https://name-resolution-exp.apps.renci.org/lookup?string=insulin&autocomplete=false&highlighting=false&offset=0&limit=10&only_taxa=NCBITaxon%3A9606&debug=none>)
- Significant improvements to documentation for Babel<https://github.com/NCATSTranslator/Babel/tree/main/docs>, NodeNorm Redis<https://github.com/NCATSTranslator/NodeNormalization#nodenormalization> and NameRes Solr<https://github.com/NCATSTranslator/NameResolution#name-resolver>
- Significant rewrite of the testing systems for these tools, especially for Babel unit tests and "pipeline tests" that exercise entire pipelines.
- Richer release notes starting with this release<https://github.com/NCATSTranslator/Babel/tree/main/releases/2026jul22> include both LLM-assisted detailed reports and multiple reports showing CURIE usage, xref used, and other information.
Please let me know if you have any questions about this release.
cheers, Gaurav
[1] https://github.com/NCATSTranslator/Babel/blob/main/releases/2026jul22 [2] https://nodenormalization-sri.renci.org/docs [3] https://name-resolution-sri.renci.org/docs