Researchers have developed an artificial intelligence model to design new superbug-fighting antibiotics.

Researchers at McMaster University and Stanford University have developed a new generative artificial intelligence model that can design billions of new antibiotic molecules that are cheap and easy to make in the laboratory.

The worldwide spread of drug-resistant bacteria has created an urgent need for new antibiotics, but modern AI methods are also limited to isolating promising chemical compounds, especially when researchers use these new AIs. Methods of making active medicines have to be found and tested. Lab

In a new study published in the journal Nature Machine Intelligenceresearchers report that they have developed a new generative AI model, called SyntheMol, that can design new antibiotics to stop the spread of Acinetobacter baumannii, which the World Health Organization has named the world’s most dangerous. It has been identified as one of the antibiotic resistant bacteria.

Notoriously difficult to eradicate, A. baumannii can cause pneumonia, meningitis and wounds, all of which can lead to death. Few treatment options remain, researchers say.

“Antibiotics are a unique drug. As soon as we start using them in the clinic, we’re starting a timer before the drugs become ineffective, because bacteria quickly evolve to become resistant to them.” ” says Jonathan Stokes, lead author of the paper and an assistant professor in McMaster’s Department of Biomedicine and Biochemistry, who worked with James Zhou, an associate professor of biomedical data science at Stanford University.

“We need a robust pipeline of antibiotics and we need to discover them quickly and cheaply. That’s where artificial intelligence plays an important role,” he says.

Researchers developed a generative model to quickly and cheaply access tens of billions of promising molecules.

They built from a library of 132,000 molecular pieces, which fit together like Lego pieces but are all very different in nature. They then cross-referenced these molecular fragments with a set of 13 chemical reactions, allowing them to identify 30 billion two-way combinations of fragments to design new molecules with the most promising antibacterial properties. can do

Each molecule designed by this model was in turn fed through another AI model trained to predict toxicity. This process yielded six molecules that exhibit potent antibacterial activity against A. baumannii and are also non-toxic.

“SyntheMol not only designs new molecules that are promising drug candidates, but it also develops the way to make each new molecule. Making these kinds of syntheses is a new approach and a game changer because chemists don’t know How AI-designed molecules are made,” says Zhou, who co-authored the paper.