TranTCR: Transformer Model for TCR-Epitope Interaction

Dependencies

See requirements.txt for all TranTCR dependencies.It is recommended to create a virtualenv and installing TranTCR within this environment to ensure proper versioning of dependencies.

Installation

You could first set up a virtual environment for TranTCR using conda:

    conda create -n TranTCR python=3.13
    conda activate TranTCR
    # Install Pytorch packages (for CUDA 12.8)
    pip3 install torch torchvision --index-url https://download.pytorch.org/whl/cu128
    pip install -r requirements.txt
    conda install -c bioconda anarci

Inference

Train

For TranTCR-bind:

    cd TranTCR-bind
    python train.py

predict

For TranTCR-bind:

    cd TranTCR-bind
    python predict.py -t ./test_data/ImmuneCode/random_mutation/unique_cdr3_1V1_Immunecode.csv -o ./output/result_test.csv

Contact

If you have any questions, please contact us at 230218235@seu.edu.cn