The abstract is read at the publisher; the summary is JClub's.
In standardized TCR-peptide-MHC testing of thousands of reported pairs, only ~50% of TCRs showed previously reported reactivity, reshaping model performance expectations.
- Why it matters: Reported TCR reactivity rates vary, leaving uncertainty about how well prediction models generalize without functional revalidation under consistent conditions.
- What they did: Using T-RAP, the study rapidly assembled large uniform TCR libraries and functionally evaluated thousands of previously reported TCR-pMHC pairs via signaling or MHC multimer binding under standardized conditions.
- The result: AlphaFold3 structural predictions identified reactive TCR-pMHC pairs within experimentally validated TCRs without task-specific training and could shortlist TCRs reactive to patient-specific cancer neoantigens.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.