Zero-shot gene and protein mention recognition in English biomedical text
Zero-shot NER (your own labels)Genes, proteins and cellsEnglishMedium · 220Mapache-2.0
What it does
This model identifies mentions of genes or proteins in English biomedical literature, including synonyms and symbol variants. It reports an F1 score of 0.799 on the BC2GM benchmark, with a 220M parameter GLiNER backbone.
Best for
- Gene-centric curation from literature
- Variant and association mining
- Custom entity recognition without retraining
What it detects (1 labels)
Trained on the label below; you can pass any other label names at inference time.
- Gene or protein
Quality
| Benchmark | BC2GM · all models on this benchmark |
|---|---|
| F1 | 0.799 |
| Precision | 0.767 |
| Recall | 0.833 |
| Accuracy | 0.946 |
| Source | Reported by the authors (test_results.json in the repository) |
Gene/Protein Entity Recognition - Gene and protein mentions
Use it
pip install gliner
from gliner import GLiNER
model = GLiNER.from_pretrained("OpenMed/OpenMed-ZeroShot-NER-Genome-Base-220M")
print(model.predict_entities("Patient with type 2 diabetes and hypertension.", ["disease"]))All model information shown here is public and belongs to its authors: use and download it freely under each model’s license. The selection, organization, search design and presentation of this catalog are © 2026 Clinical Corpus. Please link to these pages rather than copy them.