Biolexica

biolexica helps generate and apply coherent biomedical lexica. It takes care of the following:

1. Getting names and synonyms from a diverse set of inputs (ontologies, databases, custom) using pyobo, bioontologies, biosynonyms, and more.
2. Merging equivalent terms to best take advantage of different synonyms for the same term from different sources using semra.
3. Generating lexical index and doing NER using Gilda

Importantly, we pre-define lexica for several entity types that can be readily used with Gilda in the lexica/ folder including:

1. Cells and cell lines
2. Diseases, conditions, and other phenotypes
3. Anatomical terms, tissues, organ systems, etc.

Getting Started

Load a pre-defined grounder like this:

import biolexica

grounder = biolexica.load_grounder("phenotype")

>>> grounder.get_best_match("Alzheimer's disease")
Match(reference=Reference(prefix='doid', identifier='10652'), name="Alzheimer's disease", score=0.7778)

>>> grounder.annotate("Clinical trials for reducing Aβ levels in Alzheimer's disease have been controversial.")
[Annotation(text="Alzheimer's disease", start=42, end=61, match=Match(reference=Reference(prefix='doid', identifier='10652'), name="Alzheimer's disease", score=0.7339))]

Note: Biolexica constructs extended version of gilda.Grounder that has convenience functions and a more simple match data model encoded with Pydantic.

Search PubMed for abstracts and annotate them using a given grounder with:

import biolexica
from biolexica.literature import annotate_abstracts_from_search

grounder = biolexica.load_grounder("phenotype")
pubmed_query = "alzheimer's disease"
annotations = annotate_abstracts_from_search(pubmed_query, grounder=grounder, limit=30)

🚀 Installation

The most recent release can be installed from PyPI with:

pip install biolexica

The most recent code and data can be installed directly from GitHub with:

pip install git+https://github.com/biopragmatics/biolexica.git

👐 Contributing

Contributions, whether filing an issue, making a pull request, or forking, are appreciated. See CONTRIBUTING.md for more information on getting involved.

👋 Attribution

⚖️ License

The code in this package is licensed under the MIT License.

🍪 Cookiecutter

This package was created with @audreyfeldroy's cookiecutter package using @cthoyt's cookiecutter-snekpack template.

🛠️ For Developers

git clone git+https://github.com/biopragmatics/biolexica.git
cd biolexica
pip install -e .

tox

git clone git+https://github.com/biopragmatics/biolexica.git
cd biolexica
tox -e docs
open docs/build/html/index.html

tox -e finish