AI-powered RNA drug development: a new frontier in therapeutics

In the realm of modern medicine, RNA-based therapies have emerged as a promising avenue, with significant advancements in metabolic diseases, oncology, and preventive vaccines. A recent article published in Engineering titled “The Future of AI-Driven RNA Drug Development” by Yilin Yan, Tianyu Wu, Honglin Li, Yang Tang, and Feng Qian, explores how artificial intelligence (AI) can revolutionize RNA drug development, addressing current limitations and offering new opportunities for innovation.

The article highlights the potential of RNA therapies, noting that RNA drugs have shown higher success rates compared to traditional pharmaceuticals. For instance, Alnylam Pharmaceuticals claims that the cumulative transition rate of RNA interference (RNAi) drugs from clinical phase 1 to phase 3 reaches 64.4%, significantly higher than the traditional drug success rate of 5%–7%. Additionally, RNA drug discovery timelines are typically measured in months, rather than the years required for traditional drugs, and are associated with lower costs. However, despite these advantages, current experimental techniques like CRISPR and computational methods such as RNA sequencing still fall short in meeting the demands for speed and diversity in RNA drug development.

Artificial intelligence is poised to fill this gap. The article emphasizes AI’s ability to leverage parallel computing and learn complex patterns from large-scale data, thereby addressing the limitations of existing methodologies. AI-driven approaches can enhance drug development efficiency and unlock new opportunities for identifying innovative drug candidates. The authors outline three primary strategies through which AI can drive advancements in RNA drug development: data-driven approaches, learning-strategy-driven approaches, and deep-learning-driven approaches.

Data-driven approaches form the foundation by utilizing large-scale datasets and rule mining techniques to extract meaningful patterns and relationships between RNA molecules and their structures or biological functions. Learning-strategy-driven approaches employ techniques such as causal inference and reinforcement learning to optimize decision-making processes. Deep-learning-driven approaches, which represent a higher level of complexity and automation, utilize large language models like ChatGPT to analyze long RNA sequences and support the de novo design of functional RNAs.

The article envisions a future workflow for AI-driven RNA drug development that relies on an interactive, software-based system. This system would feature two key feedback loops: an internal loop focused on platform-based design to enhance AI model performance, and an external loop that integrates real-world data to continually refine drug development. The workflow would begin with comprehensive digitization of RNA data, followed by personalized drug candidate design, drug assessments, automated synthesis, and biological experiments for preliminary clinical validation. The selected drug candidates would then be matched with appropriate delivery systems and placed into an online simulation for early observation of delivery dynamics, drug action, and degradation processes within the human body.

The authors identify several challenging research topics for the near term, including high-resolution comprehensive visualization, personalized RNA drug discovery, and the development of an editable RNA generation platform. These advancements could lead to a more complete and interactive representation of RNA structures and their behavior in biological systems, enabling the creation of highly personalized RNA drugs tailored to individual genetic profiles.

The economic and social benefits of AI-driven RNA drug development are notable. AI-driven automation reduces labor-intensive tasks, enabling faster and more accurate RNA–target identification, resulting in cost savings and expedited testing of RNA therapies. As the platform scales industrially, it ensures consistent drug quality and greater cost efficiency through optimized, repeatable processes.

The integration of AI into RNA drug development holds the potential to transform the future of therapeutics. By leveraging AI’s capabilities, researchers can systematically explore novel RNA structures, identify promising drug candidates, and expedite the drug-discovery pipeline, ultimately leading to more sustainable and economical development models with widespread benefits.

The paper “The Future of AI-Driven RNA Drug Development,” is authored by Yilin Yan, Tianyu Wu, Honglin Li, Yang Tang, Feng Qian. Full text of the open access paper: https://doi.org/10.1016/j.eng.2025.06.029. For more information about Engineering, visit the website at https://www.sciencedirect.com/journal/engineering.