Machine Learning for TCR Repertoire Epitope Annotation and Pattern Discovery

Area of Science:

• Immunology
• Computational Biology
• Bioinformatics

Background:

• T cells are key to adaptive immunity, recognizing antigens via T cell receptors (TCRs).
• The vast diversity and cross-reactivity of TCRs challenge direct antigen specificity determination from repertoire sequencing.
• High-throughput sequencing generates large TCR datasets but requires computational methods to link TCRs to their target epitopes.

Purpose of the Study:

• To review and compare two main computational strategies for annotating T cell receptor (TCR) specificity: bottom-up and top-down approaches.
• To highlight the strengths and limitations of each strategy in predicting or inferring TCR-epitope interactions.
• To emphasize the importance of integrating these complementary approaches for advancing the understanding of adaptive immunity.

Main Methods:

• Bottom-up methods: Predict TCR-epitope specificity using curated databases and models (distance-based, feature-based, deep learning).
• Top-down methods: Infer antigen-driven responses from repertoire-level signals (sequence similarity, enrichment, TCR convergence).
• Review of existing literature and computational strategies for TCR specificity annotation.

Main Results:

• Bottom-up methods excel for well-characterized epitopes but suffer from data bias and poor generalization to novel epitopes.
• Top-down methods facilitate discovery of disease- or exposure-associated TCRs without prior epitope knowledge but are sensitive to noise and confounding factors.
• Both bottom-up and top-down approaches are complementary, offering mechanistic specificity and discovery capabilities, respectively.

Conclusions:

• Integrating bottom-up and top-down computational strategies is essential for accurate TCR-epitope annotation.
• Further advancements require multimodal modeling and improved benchmarking to fully understand adaptive immune responses.
• Accurate TCR specificity annotation is critical for developing targeted immunotherapies and vaccines.