Predicting T cell receptor functionality against mutant epitopes.

Area of Science:

• Immunology
• Computational Biology
• Bioinformatics

Background:

• Cancer cells and pathogens mutate epitopes to evade T cell receptors (TCRs).
• TCR cross-reactivity can counter immune evasion but risks autoimmune side effects in immunotherapies.
• Predicting T cell responses to mutated epitopes is crucial for effective cancer and infectious disease treatments.

Purpose of the Study:

• To develop a predictive model for T cell functionality against mutated epitopes.
• To assess the impact of single-point mutations on T cell receptor interactions.
• To provide a computational tool for studying T cell responses in immunotherapy.

Main Methods:

• Developed a random forest-based model named Predicting T Cell Epitope-Specific Activation against Mutant Versions (P-TEAM).
• Trained and validated P-TEAM on three datasets covering single-amino-acid mutations of model epitopes (SIINFEKL, VPSVWRSSL, NLVPMVATV).
• Evaluated model performance on 9,690 unique TCR-epitope interactions, including unseen TCRs and mutations.

Main Results:

• P-TEAM accurately classified T cell reactivities against mutated epitopes.
• The model quantitatively predicted T cell functionalities for novel single-point mutations and TCRs.
• Demonstrated high accuracy in predicting T cell responses across diverse epitope datasets.

Conclusions:

• P-TEAM is an effective computational tool for analyzing T cell responses to mutated epitopes.
• The model can aid in designing safer and more effective cell-based immunotherapies.
• Facilitates understanding of TCR-epitope interactions in immune evasion and therapeutic strategies.