New research by Prof. Yonatan Belinkov and PhD student Edo Dotan presents BetaDescribe - an innovative artificial intelligence system that translates protein sequences into natural language and provides a detailed description of their function.
The system is able to infer the function of proteins even when they are not similar to proteins already known to science, and provides information about their activity, the biological processes in which they are involved, and their binding sites - information that may aid in the development of new drugs and treatments.
The study, published in the journal PNAS, was carried out in collaboration with Prof. Tal Pupko of Tel Aviv University, and within the framework of which the researchers demonstrated the capabilities of the system by deciphering six previously uncharacterized proteins - an achievement that could significantly shorten the path between basic research and applications in medicine, biotechnology and agriculture.
BetaDescribe bridges the huge gap between the billions of proteins characterized in laboratories and the trillions of proteins existing in nature, and allows researchers to more quickly understand the role of new proteins.
Congratulations to Prof. Yonatan Belinkov, Edo Dotan and the entire research team on this impressive achievement!
Link to the article: https://www.ynetnews.com/health_science/article/skd03ulmfl
Link to the article: https://www.pnas.org/doi/abs/10.1073/pnas.2537345123
For further reading: phys.org, scienmag
[Back to the news index]