AbTools is a package of several tools for performing common antibody repertoire analysis functions:
AbCorrect corrects sequencing errors using either molecular barcodes or homology-based clustering
AbFinder identifies sequences from NGS datasets with similarity to known antibody sequences
AbPhylogeny generates publication-quality phylogenetic trees from antibody sequences
AbCompare provides methods for repertoire-level comparison of antibody sequence data
- Source code: github.com/briney/abtools
- Documentation: abtools.readthedocs.org
- Download: pypi.python.org/pypi/abtools
pip install abtools
Python 2.7 (3.x probably doesn't work, but hasn't been tested)
biopython
celery
ete2
matplotlib
numpy
pandas
seaborn
pymongo
All of the above dependencies can be installed with pip, and will be installed automatically when installing AbTools with pip. If you're new to Python, a great way to get started is to install the Anaconda Python distribution (https://www.continuum.io/downloads), which includes pip as well as a ton of useful scientific Python packages.
To calculate centroid sequences using unique antibody IDs (UAIDs) that have already been parsed by AbStar:
abcorrect -d <database_name> -c <collection_name> -t <temp_dir> -o <output_dir>
Same as above, but using a remote MongoDB server:
abcorrect -d <database_name> -c <collection_name> -i <ip> -p <port> -u <username> -p <password> -t <temp_dir> -o <output_dir>
Calculate consensus sequences using UAIDs (20 nucleotides long) that haven't been pre-parsed with AbStar:
abcorrect -d <database_name> -c <collection_name> -t <temp_dir> -o <output_dir> --parse-uaids 20 --consensus
Same as above, but UAID is 8 nucleotides long and is at the end of each sequencing read:
abcorrect -d <database_name> -c <collection_name> -t <temp_dir> -o <output_dir> --parse-uaids -8 --consensus
Calculate consensus sequences using homology-based clustering, at a threshold of 96% identity:
abcorrect -d <database_name> -c <collection_name> -t <temp_dir> -o <output_dir> --identity 0.96 --no-uaids --consensus
To generate an identity/divergence plot using known antibody sequences in a file named 'known.fasta':
abfinder -d <database_name> -c <collection_name> -t <temp_dir> -o <output_dir> -d <path/to/known.fasta>
Same as above, but without updating the MongoDB database with identity information:
abfinder -d <database_name> -c <collection_name> -t <temp_dir> -o <output_dir> -d <path/to/known.fasta> --no-update
If known.fasta contains nucleotide sequences (default is amino acid):
abfinder -d <database_name> -c <collection_name> -t <temp_dir> -o <output_dir> -d <path/to/known.fasta> --no-update --nucleotide
To compare two MongoDB collections of sequences (collection1 and collection2):
abcompare -d <database_name> -1 collection1 -2 collection2 -o <output_dir>
To iteratively compare collection1 to every other collection in the database:
abcompare -d <database_name> -1 collection1 -o <output_dir>
To iteratively compare collection1 to every other collection in the database that starts with 'abcd':
abcompare -d <database_name> -1 collection1 --collection-prefix abcd -o <output_dir>
To compare two collections (collection1 and collection2) using Jensen-Shannon similarity (default is Marisita-Horn):
abcompare -d <database_name> -1 collection1 -2 collection2 -o <output_dir> --similarity-method jensen-shannon
Generate a phylogenic tree from a FASTA-formatted file of sequences (input.fasta), a root sequence (root.fasta):
abphylogeny -i <path/to/input.fasta> -r <path/to/root.fasta> -o <output_dir>