Key points are not available for this paper at this time.
The success of COVID-19 vaccines has validated mRNA as a versatile therapeutic platform, whose expanding clinical spectrum covers immunotherapeutics for oncology (targeting oncogenes and tumour suppressor genes), autoimmunity, and infectious diseases, as well as non-immunotherapeutic applications. However, challenges in stability, delivery efficiency, and immunogenicity persist. This Review outlines how emerging RNA architectures—self-amplifying (saRNA), circular (circRNA), and trans-amplifying (taRNA)—address conventional mRNA limitations via structural engineering, while articulating the pivotal role of artificial intelligence (AI) in overcoming these hurdles across the entire mRNA drug development pipeline. We explore how AI is revolutionising key areas, such as multi-omics-guided neoantigen discovery, coding sequence and untranslated region optimisation, rational lipid nanoparticle design, and in vivo pharmacological prediction. Looking ahead, we highlight AI agents powered by genomic language models and multi-agent virtual humans (including our proposed PimRNAgent framework) that promise to automate iterative design-validation cycles and facilitate personalised mRNA therapeutics. Ultimately, the convergence of AI with next-generation mRNA platforms is paving the way for a new era of clinically actionable, precision-based mRNA therapeutics.
Tan et al. (Thu,) studied this question.