Functional evaluation of TCR-pMHC pairs at scale allows in silico TCR reactivity prediction.

Functional validation reveals only about 50% of previously reported TCR-pMHC pairs exhibit reactivity, challenging prior assumptions in TCR prediction accuracy.

• Why it matters: Accurate prediction of TCR reactivity is crucial for advancing immunotherapy and understanding immune responses, yet current models often overestimate reactivity without standardized validation.
• What they did: Using the TCR rapid assembly platform (T-RAP), the study evaluated thousands of TCR-pMHC pairs under uniform conditions and tested AlphaFold3's structural predictions without task-specific training.
• The result: Findings demonstrate that in silico methods, especially AlphaFold3, can effectively identify reactive TCRs, including those targeting cancer neoantigens, improving prospects for personalized immunotherapies.