PhD Defense: Computational methods for identifying single and combinatorial CAR T-cell targets from single-cell transcriptomics data

Chimeric antigen receptor (CAR) T cell therapy has demonstrated remarkable success in treating hematological malignancies but faces significant challenges in solid tumors due to antigen heterogeneity and on-target, off-tumor toxicities. This dissertation presents computational approaches harnessing single-cell RNA sequencing data to identify both single and combinatorial CAR cell targets with optimal safety and efficacy profiles.
Our work first established a comprehensive framework for evaluating CAR target safety and selectivity using patient tumor single-cell transcriptomics data. Through pan-cancer analysis, we demonstrated the near-optimality of existing CAR targets in most cancers while identifying novel promising targets for head and neck squamous cell carcinoma.
Building on this foundation, I developed LogiCAR Designer, a genetic algorithm that identifies logical combinations of surface antigens using Boolean operators. Applied to 17 breast cancer cohorts comprising nearly 2 million cells from 342 patients, LogiCAR Designer identified triplet antigen combinations that outperform clinically approved single-antigen targets. We further demonstrated individualized circuit design, achieving >99% tumor-targeting efficacy for most patients in a new 82-patient multi-ethnic cohort.
Finally, I applied this framework to lung cancer in non-smokers (LCINS), a critical unmet clinical need with rising incidence. Using single-cell transcriptomics data, we identified promising CAR target combinations with superior efficacy and safety profiles specific to this population.
This work provides computational tools for rational CAR target selection and precision cellular immunotherapy design, establishing a generalizable framework applicable across diverse cancer types and patient populations.