2025-268 ARCADE: Codon Optimization through AI-Driven Activation Engineering
Abstract
A codon sequence is a set of three-letter “codes” in DNA or RNA that tells cells which amino acids to link together to make proteins. mRNA (messenger RNA) is a type of molecule that carries these instructions from DNA to the cell’s protein-making machinery. Designing codon sequences carefully can change how efficiently and accurately proteins are made, which is important for areas like vaccines and gene therapy.
The invention disclosed herein is a framework for designing codon sequences with precise control over biological properties. It uses activation engineering on pretrained genomic models to adjust continuous measures such as codon adaptation index, GC content, and minimum free energy without retraining the model. This approach allows more flexible and targeted sequence design compared to traditional codon optimization methods.
Market Application
- mRNA Vaccine and Therapeutic Design: Fine-tuning codon sequences to optimize protein expression efficiency and stability for therapeutic production.
- Synthetic Biology: Generating gene sequences with desired structural and expression characteristics for engineered organisms.
- Gene Therapy Vector Optimization: Designing coding regions to improve expression in target hosts while minimizing structural instability.
- Biotech Research Tools: Providing precise control over codon-level attributes for experimental studies in molecular biology.
- Biomanufacturing: Adjusting gene constructs for large-scale recombinant protein production.
Images