Genome-wide structural modelling of TCR-pMHC interactions

Area of Science:

• Immunoinformatics
• Structural Biology
• Computational Biology

Background:

• Adaptive immunity relies on T-cell recognition of pathogen antigens via T-cell receptor-peptide-MHC (TCR-pMHC) interactions.
• Despite decades of study, the precise mechanisms of TCR-pMHC binding remain debated.
• The growing availability of structural data necessitates advanced computational tools for genome-wide modeling.

Purpose of the Study:

• To develop a fast, genome-wide structural modeling approach for TCR-pMHC interactions.
• To identify potential peptide antigens from pathogen genomes for therapeutic applications.
• To enhance understanding of immune interactions and guide peptide vaccine design.

Main Methods:

• Construction of protein-protein interaction (PPI) matrices and a novel iMatrix scoring system.
• iMatrix utilizes four knowledge-based matrices to assess hydrogen bonds and van der Waals forces.
• Inference of 701,897 potential peptide antigens from 389 pathogen genomes and modeling of TCR-pMHC complexes.

Main Results:

• The iMatrix scoring system demonstrated high correlation (Pearson's r=0.6) with experimental free energies for antigen-antibody interfaces.
• Identified peptide antigens exhibited favorable hydrogen-bond energies and consensus interactions.
• Generated TCR-pMHC models provided detailed interaction insights and highlighted crucial binding regions.

Conclusions:

• The developed method achieves high precision in predicting binding affinity and identifying potential peptide antigens.
• The iMatrix and template-based modeling approach are valuable tools for studying TCR-pMHC binding mechanisms.
• This work supports the design of novel peptide vaccines.