TIDES USA 2025: Machine Learning-Guided Ligation Fragment Design for Efficient siRNA Synthesis

The rising demand for siRNA therapeutics highlights the need for scalable and efficient manufacturing solutions. Engineered dsRNA ligases enable high-yield assembly of siRNA from short RNA fragments, with performance influenced by both ligase selection and fragment design. We introduce an AI/ML-guided approach to optimize fragment design, significantly improving ligation efficiency and enabling one-pot assembly with over 90% success. This strategy minimizes the need for extensive ligase screening, streamlining the path to scalable siRNA production.

Download the poster.