DanielKIm University of CaliforniaUnited States

Liquid biopsy is widely used for cancer detection and monitoring, but DNA-based approaches remain limited in early-stage sensitivity and dynamic biological insight. We present a nanopore-enabled RNA liquid biopsy platform that profiles full-length cell-free RNA (cfRNA) from blood which has the potential to enable early detection, longitudinal monitoring, and precision oncology applications. Long-read nanopore sequencing of samples from healthy, precancerous, and early-stage cancer research subjects reveals over 270,000 previously unannotated cfRNA transcripts, enabling construction of an expanded transcriptome reference. Machine learning enables accurate research classification of precancer and cancer while capturing pathway-level signals, including metabolic, mitochondrial, and immune checkpoint activity. This approach provides real-time, systemic readouts of tumor activity from blood, with the future potential to enable patient stratification, residual disease assessment, and monitoring of therapeutic response and emerging resistance. The platform is designed for integration with clinical workflows and DNA-based assays, supporting future, scalable deployment across diagnostic and clinical trial settings. This framework establishes a path to integrate RNA-based profiling with DNA-based approaches, potentially enhancing the impact and utility of multiomic liquid biopsies across detection, monitoring, and treatment.