● AI-Designed mRNA Increases Therapeutic Protein Expression by 15 Times

A research team led by Heo Jang, co-founder and Chief Technology Officer (CTO) of U.S. biotech venture Rayna Biosciences, revealed research results showing that AI-designed messenger RNA (mRNA) can dramatically increase protein production efficiency.

mRNA plays a role in conveying genetic information from DNA to the cellular protein synthesis machinery to instruct the production of specific proteins. mRNA was a key component in mRNA vaccines that flexibly responded to virus mutations during the COVID-19 pandemic.

The research outcome utilizing the deep learning-based generative model 'GEMORNA (Generative Models for RNA)' developed by the research team is expected to significantly enhance the effectiveness of next-generation cancer treatments and vaccines.

GEMORNA learns the entire mRNA as a language to design new sequences. Traditional mRNA design only changed some nucleotide sequences, failing to adequately reflect the complex interactions necessary for the entire mRNA to produce proteins.

GEMORNA learns genetic information as if learning a language. Just as a person considers context and grammar, not just individual words, when forming sentences, GEMORNA learns the entire mRNA sequence. It considers not only the regions directly producing proteins but also the regulatory parts before and after, designing sequences that allow cells to synthesize proteins more stably and efficiently.

Experimental results showed a significant enhancement in the expression of proteins that can be used in medical treatments. The mRNA designed by GEMORNA increased the expression of the therapeutic protein 'erythropoietin (EPO)' by 15 times compared to existing commercial sequences. The overall protein expression efficiency measured in experiments increased by up to 41 times. When a COVID-19 vaccine candidate was administered to animals, the antibody generation response was significantly enhanced, and in chimeric antigen receptor (CAR)-T cell therapy, expression stability and tumor-killing ability improved.

The research team stated that "this study is the first to prove that AI can learn and optimize the entire mRNA sequence, providing a foundation for accelerating the development of vaccines and protein therapeutics."

Popular News