On April 29, 2026, a lecture titled "AI-Based RNA Precision Therapy" was held in Zhiyuan College. The event featured Dr. Yanyi Chu from the CAS Center for Excellence in Molecular Cell Science.

Dr. Chu presented her team's latest breakthroughs in AI-driven RNA immunotherapy. To address the scarcity of high-quality non-coding RNA data, her team developed UTR-LM, a language model integrating sequence, structural, and energy information from over 500,000 experimental data points. In circular RNA research, they employed model transfer and de novo design to generate highly active IRES sequences, validated through MPRA wet-lab experiments. Furthermore, Dr. Chu discussed a Transformer-based model that predicts HLA-peptide affinity, enabling optimized neoantigen design through directed mutagenesis for RNA vaccines.

She emphasized that effective RNA drug design requires not only target selection and codon optimization but also careful consideration of complex in vivo interactions and precise expression in specific tissues.

In a supplementary session, Dr. Shaoqing Zhang explored de novo protein design, highlighting the critical role of imagination in constructing functional protein structures, from early explorations to generative design.

The interdisciplinary event drew faculty and students from several schools, including Life Sciences, Basic Medicine, and Automation. The engaging discussion covered topics such as IRES translation initiation, structural versus sequence weighting in models, and molecular network applications in RNA vaccines. The report significantly deepened attendees' understanding of RNA precision therapies and fostered enthusiasm for cross-disciplinary research.