bent 0.0.14

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 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)

closed_book:

Documentation.

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 (using Miniconda or Anaconda )

• Python>=3.7

• Required space between 5.5 GB - 10 GB

  • Dependencies: 2.5 GB

  • Data: between 3.0 GB (base) 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:

• specifying the desired knolwedge bases

• setting the argument ‘only_kb_dicts’ to True

• running the same command with these changes:

Example:

python -c "from bent.setup_package import setup_package;setup_package(['chebi'], only_kb_dicts=True)"

Reinitiate the conda environment.

Get started

To apply the complete pipeline of entity extraction (NER+NEL) set the arguments:

• recognize: indicate that the NER module will be applied (‘True’)

• link: indicate that the NEL module will be applied (‘True’)

• types: entity types to recognize and the respective target knowledge bases.

• in_dir: directory path containing the text files to be annotated (the directory must contain text files exclusively)

• out_dir: the output directory that will contain the annotation files

Example:

import bent.annotate as bt

bt.annotate(
        recognize=True,
        link=True,
        types={
         'disease': 'medic'
         'chemical': 'chebi',
         'gene': 'ncbi_gene',
         'anatomical': 'uberon',
         'cell_line': 'cellosaurus',
         'bioprocess': 'go_bp'
         },
        in_dir='input/txt/',
        out_dir='output/nel/'
)

It is also possible to apply the pipeline (NER+NEL) to a string or a list or strings instantiated in the execution script.

To see more usage examples, access the documentation.