The past several decades have seen an alarming rise in antibiotic-resistant bacteria.

Now, researchers at MIT and McMaster University in Ontario have used artificial intelligence to find an antibiotic that combats one of the worst offenders.

The research appears in the journal Nature Chemical Biology. 

Like MRSA, Acinetobacter baumannii thrives in hospitals.

To combat it, researchers “taught” a machine learning AI several molecular structures known to inhibit the germ’s growth, then had the pattern-matching algorithm look for related patterns in around 6,700 compounds.

The resulting drug, dubbed abaucin, lowers the risk of antibiotic-related C. diff because it kills the bacteria without harming beneficial gut microbes.

Better still, it ignores other harmful bacteria, thus lessening the spread antibacterial resistance.

So far this year, Arizona has reported 20 cases of A. baumannii diseases, but there’s no way to tell if other infections like pneumonia or meningitis were caused by the bacteria.