is a Data Source.
Tagger is the JensenLab dictionary-based named entity recognition (NER) engine, a fast C++ program with a Python interface that finds mentions of genes/proteins, chemicals, species, diseases, tissues, subcellular compartments, environments, phenotypes, and Gene Ontology terms in text and normalizes them to database identifiers. It is available as open-source software, as a public web service (tagger.jensenlab.org), and with precompiled dictionaries, and it produces the weekly text-mining results used by STRING, DISEASES, TISSUES, COMPARTMENTS, and ORGANISMS.
No external IDs available
| ID | Name | URL | Category | Format | Description |
|---|---|---|---|---|---|
| tagger.repository | Tagger Source Code | tagger | Process | ❔ | Source code for the tagger/tagcorpus ... |
| tagger.api | Tagger Web Service | GetEntities?document=TP53&entity_types=9606&format=tsv | Programming Interface | http | Public Tagger web service. Endpoints ... |
| tagger.dictionary | Tagger Dictionary | tagger_dictionary.tar.gz (370.7 MB) | Product | tsv | Reduced tagger dictionary used by the... |
| tagger.full-dictionary | Tagger Full Dictionary | full_dictionary.tar.gz (3.5 GB) | Product | tsv | Combined full dictionary (entities an... |
| tagger.docs | JensenLab Text-mining Resources | textmining | Documentation | http | JensenLab text-mining resources page ... |
Tagger is the dictionary-based named entity recognition (NER) engine developed in the group of Lars Juhl Jensen (Novo Nordisk Foundation Center for Protein Research, University of Copenhagen). It is written in C++ with a Python interface and produces the weekly-updated text-mining results of the JensenLab databases. Tagger expands orthographic variants (case, dashes before terminal numbers), supports curated stopword lists and entity groups (for example, orthologous proteins treated as one entity), and normalizes each match to a database identifier.
Tagger itself is software. Its outputs depend on the dictionaries it is run with. JensenLab distributes:
Gene/protein names come from the STRING alias file and chemical names from STITCH. Other entity types are built from NCBI Taxonomy, the Gene Ontology, the BRENDA Tissue Ontology, the Disease Ontology, the Environment Ontology, and phenotype ontologies. Entity types are encoded as integers: a positive NCBI Taxonomy ID means proteins of that species, and negative codes cover chemicals (-1), species (-2), GO processes, components, and functions (-21 to -23), tissues (-25), diseases (-26), environments (-27), and phenotypes.
https://tagger.jensenlab.org/GetEntities?document=...&entity_types=...&format=tsv (also GetHTML and XML output)Tagger is also used inside the EXTRACT and Reflect browser annotation tools.
The Tagger source code is released under the BSD 2-Clause License. The JensenLab website content is published under CC BY 4.0.
Created: October 19, 2026 | Last modified: October 19, 2026