Predicting immunogenicity by modeling the positive and negative selection of CD8+ T cells in individual patients
Training, prediction and evaluation can be performed by following the step-by-step instructions in the Python notebook aa.tcell_github_20221221.ipynb. The notebook was tested with Python 3.8 and TensorFlow 2.6.
The pre-trained models, training, and evaluation data can be downloaded from the following link: https://drive.google.com/file/d/1uT4zaARw6gr6j9rVXYxdxnEB9QBHmaMK/view?usp=share_link
The notebook and accompanied models and data have been organized so that one can run training/prediction/evaluation separately or run all of them from the begining. Please read the instructions in each section for more details.