GenomNER

GenomNER is a software framework for automatic recognition and tagging of genomic entities in text and sequence data. It aims to extract genes, proteins, regulatory elements, variants, and related concepts from scientific literature and genomic resources, supporting researchers in curation and data integration.

GenomNER emerged from collaboration between computational linguists and genomics researchers and is maintained as an open-source project. It provides models and tools to perform NER on domain-specific corpora and to map entities to standard identifiers in resources such as Entrez Gene and UniProt.

Techniques: The system combines tokenization, tagging, and classification using traditional methods (CRF) and neural models (BiLSTM-CRF, transformers). It can be trained on annotated corpora with gene and protein names and adapted to biomedical text, enabling cross-resource transfer.

Output and formats: GenomNER produces labeled spans with entity type, text, and cross-references to databases. It exports BIO tags, JSON, and stand-off annotations suitable for integration with genome browsers and downstream pipelines.

Availability: GenomNER is released under an open-source license and hosted on a public repository with documentation and example datasets.

Impact: The tool supports scalable annotation workflows for literature and genomic resources and has been cited in studies of automated curation.