bent 0.0.9

BENT: Biomedical Entity Annotator

Python Library for Named Entity Recognition (NER) and Linking (NEL) in the biomedical domain.

NER models are based on PubMedBERT and a post-processing rule-based module. the NEL model is a graph-based approach based on the Personalized PageRank algorithm and Information content.

BENT can be used for:

• Named Entity Recogniton (NER)

• Named Entity Linking (NEL)

• Named Entity Recognition and Linking (NER+NEL)

Access the full documentation here.

Citation:

• Pedro Ruas and Francisco M. Couto. Nilinker: attention-based approach to nil entity linking. Journal of Biomedical Informatics, 132:104137, 2022. doi: https://doi.org/10.1016/j.jbi.2022.104137.

Installation

To use the current version of BENT it is required:

• OS based on Ubuntu/Debian

• Conda environment

• Python>=3.7

• Between 11 and 15 GB free space (5 GB Anaconda + 2.5 GB to install dependencies + 3.0 GB data or 7.5 GB if you use all available knowlegde bases for NEL)

Note

Please ensure that you have the appropriate version of Conda installed.

Create a Conda environment (adapt for the name of your project):

conda create --name annotation_project python=3.7

Activate the environment:

conda activate annotation_project

Install the BENT package using pip:

pip install bent

After the pip installation, it is required a further step to install non-Python dependencies and to download the necessary data. Specify the knowledge bases that will be used:

python -c "from bent.setup_package import setup_package;setup_package([<kb1>, <kb2>, <kb3>])"

Available knowledge bases:

• ‘medic’ (MEDIC)

• ‘do’ (Disease ontology)

• ‘chebi’ (ChEBI ontology)

• ‘ctd_chem’ (CTD-Chemicals)

• ‘ncbi_gene’ (NCBI Gene)

• ‘ctd_gene’ (CTD-GENES)

• ‘ncbi_taxon’ (NCBI Taxonomy)

• ‘go_bp’ (Gene Ontology-Biological Process)

• ‘ctd_anat’ (CTD-Anatomy)

• ‘uberon’ (UBERON ontology)

• ‘go_cc’ (Gene Ontology-Cellular Component)

• ‘cell_ontology’ (Cell Ontology)

• cellosaurus’ (Cellosaurus)

Example to download only the MEDIC vocabulary:

python -c "from bent.setup_package import setup_package;setup_package(['medic'])"

If you want to download all knowledge bases, choose the option ‘all’:

python -c "from bent.setup_package import setup_package;setup_package(['all'])"

You can download more knowledge bases later by running the same command and specifying the desired knolwedge bases among the ones that are available and setting the argument ‘ only_kb_dicts’ to True:

python -c "from bent.setup_package import setup_package;setup_package([<kb>],  only_kb_dicts=True)"

Reinitiate the conda environment.