What question did this study set out to answer?

The aim is to enhance prediction accuracy of TCR recognition for peptide-MHC class I complexes via a novel framework.

February 26, 2026Open Access

TCRLens: Structure-Aware Equivariant Graph Learning for TCR-pMHC-I Recognition and Immunogenic Epitope Discovery

Key Points

The aim is to enhance prediction accuracy of TCR recognition for peptide-MHC class I complexes via a novel framework.
Developed a structure-aware deep learning framework, TCRLens, using multi-scale graph representations.
Utilized an equivariant graph neural network (EGNN) to model residue-level interactions.
Incorporated a variational autoencoder-generative adversarial network (VAE-GAN) to generate weak-affinity samples.
Evaluated TCRLens on three tasks: peptide-MHC binding, peptide-TCR recognition, and full-complex interactions.
TCRLens outperformed existing methods like NetMHCpan 4.2 and CapsNet-MHC on multiple prediction tasks.
Showed robust cross-species performance in swine and chicken MHC-I systems.
Highlighted the effectiveness of geometry-aware learning and data augmentation in improving model predictions.

Abstract

Abstract Motivation Accurate prediction of T-cell receptor (TCR) recognition of peptide-MHC class I (pMHC-I) complexes is a key challenge due to structural diversity and data sparsity. We introduce TCRLens, a structure-aware deep learning framework that models residue-level interactions across five critical interface zones using multi-scale graph representations and an equivariant graph neural network (EGNN). To mitigate data sparsity and severe class imbalance arising from limited negative samples, TCRLens incorporates a variational autoencoder-generative adversarial network (VAE-GAN) to generate structurally plausible weak-affinity interaction samples. We evaluated TCRlens across three prediction tasks including peptide-MHC binding, peptide-TCR recognition, and full-complex TCR-pMHC-I interaction and observed consistently strong performance. Results Using curated dataset of human TCR-pMHC-I structural complexes from ATLAS and TCR3d, TCRLens outperforms state-of-the-art sequence-based, motif-based, and structure-aware methods, including NetMHCpan 4.2, CapsNet-MHC, RPEMHC, NetTCR-2.0, TITAN, PanPep, pMTnet, ERGO II, and STAG. Furthermore, TCRLens demonstrates robust cross-species generalization, achieving high predictive performance in swine and chicken MHC-I systems. These findings highlight the value of geometry-aware representation learning and generative data augmentation for capturing immunological specificity. TCRLens provides a unified and extensible platform for TCR-pMHC-I interaction modeling, with potential applications in epitope discovery and structure-guided vaccine design across both human and veterinary immunology. Availability and implementation The code used in this study is publicly available at https://github.com/paopitsiri/TCRLens.

Read Full Paperexternally

Demander à l'IA

Bookmark

View Full Paper