Single Time-point Antibody Recognition
STAR (Single Time-point Antibody Recognition) is a computational pipeline designed to analyze antibody repertoires from a single time point. It enables clonal assignment and antigen specificity prediction from bulk B cell receptor (BCR) data using probabilistic modeling and machine learning.
- 🔍 Specificity Prediction: Infers antigen-specific sequences from background using statistical signals.
- 📦 OLGA Integration: Uses generative models of BCR recombination via OLGA.
Pipeline.ipynb: Main Jupyter notebook walking through the full STAR pipeline.all_class/: Custom Python classes used by the notebook.data_test/: Sample datasets for testing the pipeline.olga/: Directory containing the OLGA model used for generation probability.output/: Folder where STAR saves analysis results.
You will need:
- Python 3.6+
- Jupyter Notebook
- The following Python libraries:
numpypandasscikit-learnmatplotlibseabornolga(from https://github.com/zsethna/OLGA)
git clone https://github.com/statbiophys/STAR.git
cd STAR