Swedish AI Technology Predicts Chemical Toxicity

A groundbreaking AI method from Swedish researchers at Chalmers and the University of Gothenburg predicts chemical toxicity by analyzing molecular structure.

Share
Swedish AI Technology Predicts Chemical Toxicity

A research team from Chalmers University of Technology and the University of Gothenburg has developed a groundbreaking AI method that can effectively identify toxic chemicals. This new technology, which is both fast and cost-effective, makes it possible to assess a chemical's toxicity directly by analyzing its molecular structure.

The method uses advanced transformers to identify structural properties of molecules that contribute to toxicity. These properties can then be used to predict potential risks with the help of deep neural networks trained on extensive datasets from previous laboratory tests. These tests have involved thousands of chemicals tested on various animals and plants.

Potential for the Environment and Health

Professor Erik Kristiansson from the Department of Mathematical Sciences at Chalmers and the University of Gothenburg emphasizes the significance of the method: "There are more than 100,000 chemicals on the market today, but only a small fraction of these have established toxicity for humans and the environment." With existing methods, it isn't realistic to assess and test all of these chemicals. Kristiansson sees the new AI method as a promising alternative for tackling this challenge.

The hope is that the technology will lead to fewer laboratory tests on animals and contribute to better chemical management within both environmental research and industry. The researchers have chosen to make the method open and freely available, enabling broader use and collaboration around the world.