Metadata-Version: 2.4
Name: pybatman
Version: 1.0.1
Summary: Package for predicting peptide mutation effects on T cell receptor activation.
Project-URL: Homepage, https://github.com/meyer-lab-cshl/BATMAN
Project-URL: Bug Tracker, https://github.com/meyer-lab-cshl/BATMAN/issues
Author-email: Amitava Banerjee <amitavab@cshl.edu>
License-File: LICENSE.txt
Classifier: License :: OSI Approved :: BSD License
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Requires-Python: >=3.11
Requires-Dist: arviz>=0.21.0
Requires-Dist: numpy>=2.0.2
Requires-Dist: openpyxl>=3.1.5
Requires-Dist: pandas>=2.2.2
Requires-Dist: pymc>=5.23.0
Description-Content-Type: text/markdown

BATMAN is an interpretable Bayesian model for predicting the antigens that activate a T cell receptor (TCR). BATMAN predicts TCR activation by mutant peptides based on their distances to the TCR's index peptide. The peptide-to-index distance is a product of a learned positional weight profile vector, corresponding to effects of mutated residues at different positions in the sequence, and a learned amino acid substitution distance from the index peptide amino acid to the mutant amino acid. 

BATMAN can be trained in two modes: (1) within-TCR, where the train and test peptides are associated with the same TCR, and BATMAN-inferred positional weight profiles are TCR-specific, and (2) leave-one-TCR-out, where peptides are tested for activation of a TCR left out of the training data, and BATMAN-inferred positional weight profile is common across all TCRs. 

BATMAN outperforms existing TCR-antigen methods, reveals structural and biochemical predictors of TCR-antigen interactions, and can predict polyclonal T cell responses and TCR targets with high sequence dissimilarity. For installation, usage, tutorials and more information, refer to https://github.com/meyer-lab-cshl/BATMAN.