AI can quickly identify structure of drugs suitable for ‘legal highs’

An AI tool can easily suggest possible individuals for the chemical structures of psychoactive “designer drugs” from a simple analysis. The tool could fast-track the development of tests that screen the utilization of drugs with similar effects to substances such as for example cocaine and heroin, but aren’t detectable with current tests.

“Our method could decrease the time required to identify a fresh designer drug from weeks or months to just hours,” says Michael Skinnider at the University of British Columbia in Canada.

Skinnider and his colleagues created a machine learning tool called DarkNPS by training it with chemical structures of around 1700 known designer drugs, collected from forensic labs all over the world. Working out set included tandem mass spectrometry results for every drug, that is a common technique that provides information on the mass of a molecule and the factors it includes. This allowed the AI to recognize patterns between tandem mass spectrometry data and chemical structures.


Read more: Opioid levels in US wastewater spiked through the early pandemic months

Given tandem mass spectrometry data for a previously unseen drug, DarkNPS could then guess the molecular structure with an accuracy of 51 %. This number increased to 86 % if the AI could give its top 10 10 predictions of its structure, meaning that it could be most useful for narrowing the search.

“This may save an enormous amount of time and be able to recognize new designer drugs much sooner after they’ve to enter the market,” says Skinnider.

The researchers also used the tool to look at drugs that could be created later on utilizing the AI to generate 1 billion possible chemical structures. Afterwards, the team acquired data for 194 new designer drugs and found that 176 of the appeared in the set produced by the AI.

Journal reference: Nature machine intelligence , DOI: 10.1038/s42256-021-00407-x

More on these topics:

  • artificial intelligence
  • drugs

Related Posts

Leave a Reply

Your email address will not be published.