Cambridge, Mass.-based MIT and Ontario, Canada-based McMaster University researchers have found a new antibiotic treatment that can kill a common bacteria in hospital infections thanks to machine learning.
The study, published May 25 in Nature Chemical Biology, used machine learning to identify which chemicals could inhibit the growth of Acinetobacter baumannii, a common hospital bacterium. After analyzing 6,680 compounds in two hours, the algorithm identified a few hundred options. Researchers chose 240 to test in the lab, which yielded nine antibiotics. One compound, originally explored as a potential diabetes drug, was effective in killing A. baumannii but had no effect on other species of bacteria — a desirable attribute.
In mice studies, researchers showed the drug, which they named abaucin, treated wound infections caused by A. baumannii, according to a May 25 MIT news release. In lab tests, it also worked against a variety of drug-resistant A. baumannii strains isolated from human patients.
"This finding further supports the premise that AI can significantly accelerate and expand our search for novel antibiotics," James Collins, PhD, professor of medical engineering and science in MIT's Institute for Medical Engineering and Science and department of biological engineering, said in the release. "I'm excited that this work shows that we can use AI to help combat problematic pathogens such as A. baumannii."