Artificial Intelligence Finds an Antibiotic

In a news flash from M.I.T. scientists announce the discover of a (potentially) low-toxicity molecule which interferes with bacteriological cell walls’ ability to maintain electrochemical gradients, which are essential to creating ATP, the main energy molecule in, well, everything: the cells would starve. An A.I. was trained on 2,500 molecules and then scanned the Broad Institute’s Drug Repurposing Hub, a library of about 6,000 compounds. The model picked out one molecule that was predicted to have strong antibacterial activity and had a chemical structure different from any existing antibiotics. Using a different machine-learning model, the researchers also showed that this molecule would likely have low toxicity to human cells.

It worked very well in vitro and in mouse models on a bunch of stubborn microbes that are getting to be pretty resistant to everything we have: Clostridium difficile, Acinetobacter baumannii, and Mycobacterium tuberculosis. The drug apparently worked on EVERYTHING they tested, except Pseudomonas aeruginosa.