Scientists have discovered a new antibiotic that kills treatment-resistant diseases, using artificial intelligence.

The new drug, named halicin, was found using a computer model that could pave the way for other powerful antibiotics.

Its discovery was reported by researchers from Massachusetts Institute of Technology (MIT) in a study published in the journal Cell.

Halicin, which had previously been explored as a diabetes drug, successfully killed many antibiotic-resistant bacteria including Clostridium difficile, Acinetobacter baumannii and Mycobacterium tuberculosis.

‘In terms of antibiotic discovery, this is absolutely a first,’ said Regina Barzilay, a senior researcher on the project at MIT.

According to the researchers, the computer model used to discover halicin has the potential to find hundreds of other powerful antibiotics and has already identified 23 promising candidates. 

Professor of biology and computer science at Technion in Israel, Roy Kishomy, added: ‘This groundbreaking work signifies a paradigm shift in antibiotic discovery and indeed in drug discovery more generally.’ 

In order to discover halicin, researchers used a machine learning model, where they trained a computer to identify the sorts of molecules that kill bacteria. 

Once the model had learnt which molecules were good at killing bacteria, the scientists set it working on a library of more than 6,000 compounds. It selected one molecule that it predicted would be successful at killing bacteria. 

The researchers now plan to study whether halicin can be developed for human use.