AI Helps Discover Promising New Antibiotics to Combat Drug-Resistant Infections
-
Scientists discovered a new class of antibiotics effective against drug-resistant Staphylococcus aureus (MRSA) using AI models to predict antibiotic activity and toxicity.
-
The models screened 12 million commercially available compounds and identified 5 classes of molecules predicted to combat MRSA.
-
Tests of 280 compounds from those classes in the lab yielded 2 promising new antibiotic candidates that reduced MRSA levels in mouse infection models.
-
The AI models were more transparent deep learning models that allowed researchers to better understand what features led to accurate predictions.
-
This discovery of a compound that can kill this deadly drug-resistant bacterium could be a major breakthrough in fighting antibiotic resistance.