In a report published July 25 in Nature, a group of UT researchers created a novel artificial intelligence model that can help with drug and vaccine discovery by forecasting particular genetic patterns. These sequences have the ability to generate essential proteins that activate the immune system.
The concept, known as RiboNN, was created in collaboration with the global pharmaceutical and healthcare corporation Sanofi. Can Cenik, an associate professor of molecular biosciences and one of the model’s co-researchers, stated that when his team started developing the model six years ago, they had no intention of including artificial intelligence. According to him, the usage of AI has only happened in the last three years of the study.
According to Cenik, the process by which messenger RNA, or mRNA, is taken and transformed into proteins is known as translation. According to him, this process takes place in every cell and is necessary for life. Additionally, it may result in the development of mRNA vaccines, according to MedLinePlus.
According to MedLinePlus, the body can be instructed to produce particular proteins that aid in the fight against an infection via the genetic code carried by mRNA. On the outside of their cellular membranes, viruses have a unique protein. The same protein that the virus creates is produced by the code once it enters the body. In order to prevent further infection, the body’s immune system reacts to these alien proteins by producing particular antibodies that multiply and kill them.
By using a data source developed by the research lab, the team’s model can successfully anticipate the code for these particular mRNA sequences, according to Cenik. He stated that testing on human and mouse cell types are part of the team’s mRNA production research, which is included in the data source. According to the news release, the model may help researchers speed up therapies for infectious diseases and other types of cancer. One possible application is to eventually support therapeutic interventions aimed at specific cells.
Cenik stated that since he has already done this using conventional computers or machine learning, he does not believe that RiboNN represents a significant change from conventional methods of mRNA sequence prediction.
Cenik stated, “I have been working in machine learning, which is essentially using computers to make predictions about things, for more than 20 years.” This is merely the following generation.
This technology was made publicly available for use by other research groups by Cenik and his team. He expects that others will use it to further scientific discoveries.
According to Cenik, it’s always amazing to witness others using your work in inventive ways you never would have thought possible. When other people utilize it for their studies, that’s actually where it makes the largest difference. I find that to be really satisfying.
