Artificial intelligence helped identify a promising new antibiotic. Researchers deployed AI to zero in on a chemical that might be the nemesis of the notorious Acinetobacter baumannii. The CDC reports that this drug-resistant “superbug” caused 8,500 infections and 700 deaths in hospitals during 2017 alone.
In this pioneering work, they showed a potential antibiotic called abaucin had the potential to stop the growth of this stubborn bacterium.
They developed an AI model to determine which of 7684 chemical compounds would be effective against the bacterium. In a couple of hours, their model identified a molecule they named abaucin as a potential antibiotic. Without AI, this research might have taken months of work. These results, of course, need further evaluation through larger scale studies.
The researchers call the use of AI to spotlight promising drug candidates “incredibly promising.” This approach strategy might reshape drug discovery research.
Jonathan Stokes, the study’s lead author, views AI models as tools that can enhance our efficiency. “Think of these AI models as our allies, making us better at what we do,” he says.
Current screening can evaluate a few million drugs or chemical components simultaneously. AI-powered algorithms, however, can assess hundreds of millions, or even billions, of drug molecules at a time. Cesar Fuente, a medical researcher from the University of Pennsylvania, called this new research method “incredibly promising” and predicted that drug discovery is the next frontier.
1 comment
Comments are closed.