| 研究実績の概要 |
We have achieved the main objective of this project: to develop a general method to predict the epitope targeted by a given T cell receptor (TCR) and major histocompatibility complex (MHC). To build a 3D model of each TCR sequence, we have developed Repertoire Builder, which can generate 10,000 antibody or TCR models in approximately 30 minutes more accurate than any other tested server (Schritt, D. et al. MSDE 2019). Next, to cluster the TCRs, we developed InterClone, which can cluster antibodies or TCRs into epitope-specific groups (Xu, Z. et al. MSDE 2019). Next, we can construct a 3D model of the TCR-epitope-MHC complex using ImmuneScape (Li, S. et al. Meth Mol Biol 2019). The final step in the pipeline is to develop a score that can distinguish true peptide epitopes from false ones. We have developed such a score and demonstrate that it dramatically out-performs the original ImmuneScape score. Using the above pipeline, we have made significant contributions to the identification of antigens involved in allergy (Takeda, K. et al. J Allergy Clin Immunol 2019), modeled antibodies involved in enhancing SARS-CoV-2 infection (Liu, Y. et al. Cell 2021), modeled CAR T cell receptors (Singh, N. et al Nat Med 2021) and modeled the self-reactive epitopes involved in neuromyelitis optica spectrum disorder (in prep). These tools are currently being applied to the identification of SARS-CoV-2 T cell epitopes in Japanese COVID-19 patients.
|
