trimap-tools is a package for analysis of peptide-HLA presentation and TCR specificity. It is designed to help researchers understand the interactions between T cell receptors (TCRs) and peptides presented by human leukocyte antigen (HLA) molecules, which play a crucial role in the immune response.
Create a conda environment. trimap-tools requires Python 3.9 or later.
conda create -n trimap-env python=3.9
conda activate trimap-envIt is available on PyPI and can be installed using pip:
pip install trimap-toolsor by cloning the repository and installing it manually:
pip install git+https://github.com/uhlerlab/trimap-tools.git@mainFor step-by-step guides on how to use trimap-tools, including training HLA/peptide encoders and predicting TCR specificity, please refer to the Documentation section.
To predict TCR specificity for pHLA complexes, follow these steps:
-
Train representations for HLA sequences or load our pre-trained models from Here.
-
Train representations for peptides in the context of HLA or load our pre-trained models from Here.
-
Predict TCR specificity for pHLA complexes or load our pre-trained models from Here.
To train a model for discovering disease-associated epitopes, follow these steps:
-
Peptide representation learning with HLA context
Learns latent embeddings of peptides while incorporating HLA background, enabling biologically informed modeling. -
TCR specificity prediction using full receptor sequences
Supports comprehensive modeling of TCR recognition by leveraging both α and β chain CDR regions. -
Visualization of critical TCR residues
Highlights key amino acid positions in TCRs that contribute to antigen recognition, aiding biological interpretation. -
Discovery of disease-associated epitopes
Identifies novel peptides potentially involved in disease by integrating small-scale disease-specific data with large-scale public datasets.
If you use trimap-tools in your research, please cite the following paper: