Introduction Identifying neoantigen-specific T-cell receptor (neoTCR) clonotypes is crucial for tracking clinical T-cell dynamics in personalized cancer immunotherapy. Despite advances in experimental and computational approaches for identifying antigen-specific TCR (asTCR) clonotypes, the identification of neoTCR clonotypes remains challenging due to their low frequencies and the limited sensitivity of current methods. Methods We introduce NeoTCRseek, an integrated platform that combines extended T-cell culture, deep TCR sequencing, and advanced TCR-clustering tools to enhance neoTCR clonotype identification. NeoTCRseek was developed using a model cytomegalovirus (CMV) antigen and subsequently validated under two neoantigen setups: a single neoantigen and a neoantigen pool representing a multi-antigen context. For each antigen setup, we used three cell sorting-based methods to detect enriched clonotypes and built a validation dataset for asTCR clonotype characterization and NeoTCRseek performance evaluation. Results NeoTCR clonotypes exhibited a significantly higher proportion of low-frequency clonotypes (0.01%–0.1%) than CMV-specific ones (70.3% vs. 33.3%). Nonetheless, NeoTCRseek achieved a detection limit of 0.01% and high accuracy in both single- and multi-antigen contexts by integrating expanded clonotype detection with co-clustering-based TCR prediction. Compared with the benchmark, NeoTCRseek improved the mean F1 score across the two neoantigen setups from 0.21 to 0.41. Conclusion NeoTCRseek achieves high analytical sensitivity and accuracy in neoTCR clonotype identification and supports multi-antigen analysis, providing an integrated platform for neoTCR clonotype characterization and for tracking clinical T-cell dynamics in personalized cancer immunotherapy.
Song et al. (Wed,) studied this question.