TCRconv

TCRconv predicts interactions between T cell receptors (TCRs) and epitopes presented by major histocompatibility complexes to model antigen recognition and study T cell-mediated immune responses.

Key Features:

• Deep Learning Model: Employs a deep protein language model combined with convolutional techniques to extract contextualized motifs from TCR and epitope sequences.
• High Prediction Accuracy: Integrates advanced machine learning methodologies to achieve high prediction performance of TCR-epitope recognition compared to existing models.
• T cell Repertoire Analysis: Analyzes T cell receptor repertoires, including data from COVID-19 patients, to provide insights into T cell dynamics and phenotypes.
• Research Applications: Applies predictive mappings of TCR-epitope interactions to support immunotherapy development, vaccine design, and mapping immunological signatures for personalized medicine and disease studies.

Scientific Applications:

• Infectious disease research (COVID-19): Characterizes T cell responses and repertoire changes in COVID-19 patients to study disease-associated immune signatures.
• Autoimmunity: Identifies TCR-epitope recognition patterns relevant to autoimmune disease-associated antigens.
• Tumor immunology: Maps potential tumor-associated epitope recognition by TCRs to inform tumor immune profiling.
• Immunotherapy and vaccine development: Supports prediction of antigen-specific TCRs for designing immunotherapies and vaccines.
• Personalized medicine: Enables mapping of individual immunological signatures based on TCR-epitope interaction predictions.

Methodology:

Applies a deep protein language model to TCR and epitope sequences and uses convolutions to capture contextualized sequence motifs for predicting TCR-epitope recognition.