AI-integrated RNA therapeutics in precision oncology.

Abstract

RNA therapeutics can silence oncogenic transcripts, modulate RNA processing, restore protein expression, or encode tumor antigens. Their development, however, is constrained by target selection, molecular design, delivery, immunogenicity, and inter-patient heterogeneity. Artificial intelligence (AI) is increasingly applied to these bottlenecks through multi-omics target prioritization, RNA structure and sequence modeling, delivery optimization, neoantigen selection, patient stratification, and manufacturing quality control. The maturity of these applications is uneven: within oncology, messenger RNA (mRNA) cancer vaccines are the most clinically advanced RNA-based strategy, whereas therapeutic approaches involving long non-coding RNAs and circular RNAs remain predominantly preclinical or early clinical, and most AI-enabled methods lack prospective clinical validation. In this review, we evaluate AI across the RNA therapeutic development pipeline, distinguish evidence-supported applications from preclinical and speculative uses, and discuss the validation, regulatory, and data-governance requirements for translation into precision oncology.