This repository contains the models and predictions used for the NetTCR entries for the IMMREP 2023 Kaggle competition.
First, ensure that miniconda/anaconda is installed (see https://docs.conda.io/projects/miniconda/en/latest/ for more info).
Following this, the conda environment for NetTCR-2.2 can be installed by running conda env create -f environment.yml from the NetTCR-2.2 folder. This will create a conda environment called nettcr_2_2_env with the necessary dependencies.
To load this environment afterwards in commandline, run the command conda activate nettcr_2_2_env.
NetTCR-2.2 is freely available for academic user for non-commercial purposes (see license). The product is provided free of charge, and, therefore, on an "as is" basis, without warranty of any kind. Other users: If you plan to use NetTCR-2.2 or any data provided with the script in any for-profit application, you are required to obtain a separate license. To do so, please contact health-software@dtu.dk.
For licence details refer to academic_software_license_agreement.pdf
Jensen, M. F., & Nielsen, M. (2023). NetTCR 2.2—Improved TCR specificity predictions by combining pan- and peptide-specific training strategies, loss-scaling and integration of sequence similarity. bioRxiv. https://doi.org/10.1101/2023.10.12.562001
Justin Barton. (2023). IMMREP23: TCR Specificity Prediction Challenge. Kaggle. https://kaggle.com/competitions/tcr-specificity-prediction-challenge