Abstract LB189: Prospective AI/ML-driven computational approaches for predicting T cell-mediated immunity and beyond

Abstract Background: T cells play a double-edged sword role in various physiopathological states. They have the ability to eliminate tumors and infected cells, and play an essential role in host defense. Instead, T cells can play pathogenic roles, activated by autoantigens and causing damage to self-tissues. Therefore, triggering and fine-tuning the control of T cell-peptide/major histocompatibility complex (MHC) (TCR-pMHC) recognition is crucial for maintaining the balance of many aspects of disease and pathological states for the treatment of cancer, infectious, and autoimmune disease. Methods: Despite the previous interest and attempt, generalizable predictive models of TCR-pMHC specificity remain nascent. Inspired by recent breakthroughs in protein structure prediction achieved by deep neural networks, we evaluated structural modeling using AlphaFold 3 (AF3)-based AI-enabled computation as a potential avenue for predicting TCR epitope specificity. Results: We show that a specialized version of the neural network predictor AlphaFold can generate models of TCR-pMHC interactions that can distinguish valid from invalid peptide epitopes with increasing accuracy. Accurate prediction of T-cell immunogenicity holds significant therapeutic value. It helps identify peptide epitopes linked to tumors, infectious agents, and autoimmune diseases. Higher-affinity T cells can be computationally designed with a higher success rate for T cell therapy. Strong immunogenicity of antigenic epitopes can be selected for vaccine development. Conclusions: An accurate TCR-pMHC model will significantly benefit T-cell-mediated immunotherapy and facilitate advanced drug design. Although much work remains before these predictions achieve widespread practical utility, we are optimistic that deep learning-based structural modeling represents a path to the generalizable prediction of TCR-pMHC interaction and beyond. Citation Format: Chongming Jiang, Yulun Chiu, Xiling Shen, Cheng-chi Chao. Prospective AI/ML-driven computational approaches for predicting T cell-mediated immunity and beyond [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr LB189.