Full text is available at the source.
Abstract
RNASTOP achieves a 13% improvement in accuracy for predicting mRNA chemical stability compared to the top-performing model.
- RNASTOP integrates deep learning with heuristic search to enhance mRNA stability predictions.
- The framework demonstrates robust generalization in predicting full-length mRNA degradation.
- When optimizing mRNA sequences, RNASTOP reduces the minimum free energy of a varicella-zoster virus vaccine sequence by 75.73%.
- High translation efficiency is maintained during the optimization process.
- RNASTOP could serve as a valuable tool for advancing mRNA therapeutic development.
Simplified