PiTE: Official Tensorflow Implementation

We provide the official Tensorflow & Keras implementation of training our PiTE: TCR-epitope binding affinity prediction pipeline using Transformer-based Sequence Encoder.

Publication

PiTE: TCR-epitope Binding Affinity Prediction Pipeline using Transformer-based Sequence Encoder
Pengfei Zhang^1,2, Seojin Bang², Heewook Lee^1,2
1 School of Computing and Augmented Intelligence, Arizona State University, ² Biodesign Institute, Arizona State University
Published in: Pacific Symposium on Biocomputing (PSB), 2022.

Paper | Code | Poster | Slides | Presentation (YouTube)

Dependencies

• Linux
• Python 3.6.13
• Keras 2.6.0
• TensorFlow 2.6.0

Usage of PiTE

1. Clone the repository

git clone https://github.com/Lee-CBG/PiTE
cd PiTE/
pip install -r requirements.txt

2. Download training and testing dataset

• Data for baseline model can be downloaded here. The size is 5.16 GB.
• Data for other models such as Transformer, BiLSTM, and CNNs can be download here. The size is 68.93 GB. Preprocess this data using preprocessing.ipynb file before intiating the training.

3. Train models

An example for training the transformer-based model

python -W ignore main.py \
--nns transformer \
--split tcr \
--gpu 0 \
--run 0 \
--seed 42

Citation

If you use this code or use our PiTE for your research, please cite our paper:

@inproceedings{zhang2022pite,
  title={PiTE: TCR-epitope Binding Affinity Prediction Pipeline using Transformer-based Sequence Encoder},
  author={Zhang, Pengfei and Bang, Seojin and Lee, Heewook},
  booktitle={PACIFIC SYMPOSIUM ON BIOCOMPUTING 2023: Kohala Coast, Hawaii, USA, 3--7 January 2023},
  pages={347--358},
  year={2022},
  organization={World Scientific}
}

License

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.