EUVIMEDEUVIMED entdecken
Crossref

ImmunoFoundation: A Multimodal Deep Learning Approach to Immunogenicity Prediction 2310036

Smita Krishnaswamy, João Felipe Rocha, Hiren Madhu, Rishabh Anand, Chen Liu, Apurva Mishra, Akiko Iwasaki

The Journal of Immunology · 01.08.2026 · 215 (Supplement_1)

DOI: 10.1093/jimmun/vkag141.1676 ↗

Kurz gefasst

Fachartikel in der Zeitschrift The Journal of Immunology, online erschienen am 28.07.2026. Verfasst von Smita Krishnaswamy und 6 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

Predicting immunogenicity remains a critical challenge in vaccine design, autoimmunity treatment, and pharmaceutical development. Current AI tools rely on limited data inputs, typically only class-I MHC-peptide sequences, missing crucial structural and biochemical information. We developed the ImmunoFoundation Model (IFM), a multimodal deep learning system that integrates not only peptide sequences, 3D molecular structures, and biochemical properties but also TCR-MHC-peptide (class-II and class-I) to achieve superior immunogenicity prediction and enable peptide optimization for therapeutic applications.

Methods

IFM comprises three modules: (1) ESM3 transformer for embedding antigen-peptide, MHC, and TCR sequences from vast biomedical data; (2) geometric scattering transformer networks to capture molecular structure from AlphaFold3-predicted peptide-MHC complexes; (3) Autoencoder for biochemical property embedding including surface area and thermal stability. Cross-modal attention layers integrate these representations. Training utilized IEDB, VDJdb, McPAS-TCR, and TCR3d datasets totaling >300,000 samples across MHC class I and II.

Results

The preliminary model achieved state-of-the-art performance on CEDAR cancer neoepitope datasets. Attention mechanism analysis revealed structural motifs influencing immunogenicity, distinguishing between KRAS G12V and G12D mutants. The model successfully predicted vaccine cassette immunogenicity and identified key peptide-MHC interaction sites. Current IFM development shows improved multimodal integration with enhanced predictive accuracy across viral and cancer peptide immunogenicity tasks.

Conclusion

IFM represents a paradigm shift in immunogenicity prediction by comprehensively modeling the complex antigen-MHC-TCR interaction system. Its generative capabilities enable peptide optimization for cancer vaccines and personalized immunotherapy, with potential applications in autoimmunity treatment and biologics development. Funding Source Yale Colton Center for Autoimmunity Topic Categories Computational and Systems Immunology (COMP)

Zitieren

APA

Krishnaswamy, S., Rocha, J. F., Madhu, H., Anand, R., Liu, C., Mishra, A., & Iwasaki, A. (2026). ImmunoFoundation: A Multimodal Deep Learning Approach to Immunogenicity Prediction 2310036. The Journal of Immunology, 215(Supplement_1). https://doi.org/10.1093/jimmun/vkag141.1676

Vancouver

Krishnaswamy S, Rocha JF, Madhu H, Anand R, Liu C, Mishra A, et al. ImmunoFoundation: A Multimodal Deep Learning Approach to Immunogenicity Prediction 2310036. The Journal of Immunology. 2026;215(Supplement_1). doi:10.1093/jimmun/vkag141.1676

BibTeX
@article{Krishnaswamy2026,
  author = {Krishnaswamy, Smita and Rocha, João Felipe and Madhu, Hiren and Anand, Rishabh and Liu, Chen and Mishra, Apurva and Iwasaki, Akiko},
  title = {ImmunoFoundation: A Multimodal Deep Learning Approach to Immunogenicity Prediction 2310036},
  journal = {The Journal of Immunology},
  year = {2026},
  volume = {215},
  number = {Supplement_1},
  doi = {10.1093/jimmun/vkag141.1676}
}
Kostenlos starten →