Related Experiment Video
Updated: Mar 17, 2026
16:10
A TIRF Microscopy Technique for Real-time, Simultaneous Imaging of the TCR and its Associated Signaling Proteins
Published on: March 22, 2012
24.5K
Decoding TCR recognition via geometric deep learning of immunological fingerprints.
Chun Shang1,2, Kevin C Chan3, Ruhong Zhou1,2,4,5
1College of Physics, College of Life Sciences, and Institute of Quantitative Biology, Zhejiang University, 866 Yuhangtang Road, Hangzhou 310058, China.
Briefings in Bioinformatics
|March 16, 2026
Summary
A new deep learning model decodes T cell receptor (TCR) recognition by predicting peptide-MHC binding. This framework reveals "immunological fingerprints" to understand immune responses and guide therapies.
Area of Science:
- Immunology
- Structural Biology
- Computational Biology
Background:
- T cell receptor (TCR) recognition of peptide-major histocompatibility complex (pMHC) is crucial for adaptive immunity, pathogen defense, and self-tolerance.
- Understanding TCR specificity and cross-reactivity remains challenging despite extensive structural data.
- Existing structural analyses struggle to capture the complex features governing TCR-pMHC interactions.
Purpose of the Study:
- To develop a deep learning framework for predicting TCR recognition of pMHC molecules.
- To identify key structural and physicochemical features driving TCR-pMHC engagement.
- To provide interpretable insights into TCR specificity and cross-reactivity.
Main Methods:
- Developed a multimodal geometric deep learning framework to extract spatial and physicochemical features from pMHC interfaces.
- Applied the model to a dataset of HLA-A*02-peptide-TCR crystal structures for training and validation.
- Integrated an explainability module to identify critical interaction residues and motifs.
Main Results:
- The deep learning model accurately predicts TCR binding preferences.
- Identified unique interfacial "immunological fingerprints" that dictate TCR recognition.
- Revealed potential TCR cross-reactivity between self and bacterial peptides using HLA-B*27 complexes.
Conclusions:
- Established a scalable, structure-based deep learning approach for decoding T cell recognition.
- The framework offers interpretable insights into TCR specificity determinants.
- Provides a powerful tool for antigen design, vaccine development, and TCR-based immunotherapies.
