abutils

Models, functions and visualization tools for working with adaptive immune receptore repertoire (AIRR) data. The primary purpose of abutils is to provide generalizable tools suitable for direct use analyzing bulk AIRR datasets, and is used by scab for single cell AIRR analysis. abutils is a core component of the ab[x] toolkit for AIRR data analysis.

• Source code: github.com/briney/abutils
• Documentation: abutils.readthedocs.org
• Download: pypi.python.org/pypi/abutils
• Docker: hub.docker.com/r/brineylab/datascience/

install

pip install abutils

api

We've tried to design the abutils API to be intuitive yet powerful, with the goal of enabling both interactive analyses (via environments like Jupyter notebooks) as well as integration of abutils tools into more complex analysis pipelines and/or standalone software tools. See the documentation for more detail about the API. As always, any feedback is greatly appreciated!!

testing

You can run the complete abutils test suite by first installing pytest:

pip install pytest

and then running:

git clone https://github.com/brineylab/abutils
cd abutils
pytest

This test suite is automatically run against all supported versions of Python following every commit.

requirements

python 3.10+

abstar
baltic
biopython
dnachisel
fastcluster
matplotlib
mnemonic
natsort
numpy
pandas
parasail
polars
prettytable
pyarrow
pyfamsa
pyfastx
pytest
python-circos
pyyaml
rapidfuzz
sample-sheet
scikit-learn
scipy
seaborn
smart_open
tqdm

abutils includes several additional binaries that are required for certain functionality:

• abutils.tl.mafft uses MAFFT
• abutils.tl.muscle uses MUSCLE
• abutils.tl.cluster uses CD-HIT, MMseqs2, and VSEARCH
• abutils.tl.fasttree uses FastTree

Although these binaries are all packaged into abutils, each respective abutils.tl function provides the option to supply a alternate binary path in case you'd prefer to use a different version.