Researchers in the US have used a high-powered artificial intelligence (AI) model to discover new families of antibiotics called "halicin" that can kill drug-resistant superbugs.

"We wanted to develop a platform that would allow us to harness the power of artificial intelligence to usher in a new age of antibiotic drug discovery," says James Collins, senior author of the study that appears in the journal Cell.  "Our approach revealed this amazing molecule which is arguably one of the more powerful antibiotics that has been discovered."

Ever since antibiotics were first discovered, modern medicine has been in an arms race with dangerous bacteria which are developing resistance to drugs much faster than new ones can be developed.  It takes crunching through enormous amounts of data to find molecules with the right attributes - luckily, that's exactly what AI excels at.

The team made of scientists from the Massachusetts Institute of Technology (MIT) and Harvard loaded up data on 2,500 molecules and 6,000 drug compounds, and sent the AI on a mission to find the ones that can fight a familiar menace.

"In this case, the researchers designed their model to look for chemical features that make molecules effective at killing E. coli," MIT wrote in a statement.  "To do so, they trained the model on about 2,500 molecules, including about 1,700 FDA-approved drugs and a set of 800 natural products with diverse structures and a wide range of bioactivities."

Halicin worked on the following bacteria in both petri dish tests and in laboratory mice: Clostridium difficile, known as C Diff; Acinetobacter baumannii, which killed 700 people in 2017 and has vexed allied troops stationed in Iraq and Afghanistan; and Mycobacterium tuberculosis, which causes TB or "consumption".

In addition to their ongoing studies of halicin, the team used the AI-discovered model to uncover 23 other candidate molecules that could have antibiotic abilities as well.