New AI method can detect toxic chemicals

Chemicals are widely used in society and are found in everything from household products to industrial processes. Many chemicals end up in our waterways and ecosystems, where they can have negative effects on humans and other organisms. One example of this is PFAS, a group of problematic substances that have recently been found in worrying concentrations in both groundwater and drinking water. They have been used, for example, in fire-fighting foam and in many consumer products.

Negative effects on humans and the environment occur despite extensive chemical regulations, which often require time-consuming animal testing to prove when chemicals can be considered safe. In the EU alone, more than two million animals are used every year to comply with the various regulations. At the same time, new chemicals are being developed at a rapid pace and it is a major challenge to determine which ones need to be restricted due to their toxicity to humans and the environment.

Valuable help with the development of chemicals

The new method developed by the Swedish researchers uses artificial intelligence to assess chemical toxicity quickly and cost-effectively. It can therefore be used to identify toxic substances at an early stage and reduce the need for animal testing.

"Our method is able to predict whether a substance is toxic or not based on its chemical structure. It has been developed and refined by analyzing large data sets of laboratory tests that have been carried out in the past. The method has been trained to give accurate estimates even for previously untested chemicals," says Mikael Gustavsson, researcher at the Department of Mathematical Sciences at Chalmers University of Technology and the Department of Biology and Environmental Sciences at the University of Gothenburg.

"There are currently more than 100,000 chemicals on the market, but only a small proportion of these have a well-described toxicity to humans and the environment. Assessing the toxicity of all these chemicals using conventional methods, including animal testing, is practically impossible. This shows that our method can offer a new alternative," says Erik Kristiansson, professor at the Faculty of Mathematical Sciences in Chalmers and at the University of Gothenburg.

The researchers believe that the method can be very useful for environmental research as well as for authorities and companies that use or develop new chemicals. They have therefore made publicly accessible.

More comprehensive and accurate than today's calculation tools

Computerized tools for detecting toxic chemicals already exist, but until now their applications have been too narrow or their accuracy too low to replace laboratory testing on a large scale. In the researchers' study, they compared their method with three other commonly used computer-based tools and found that the new method is both more accurate and more widely applicable.

"The type of AI we use is based on advanced deep learning methods," says Erik Kristiansson. "Our results show that AI-based methods can already keep up with conventional computational approaches, and as the amount of available data continues to increase, we expect AI methods to improve further. Therefore, we believe that AI has the potential to significantly improve the computational assessment of chemical toxicity."

The researchers predict that AI systems will be able to replace laboratory tests to an ever greater extent.

"This would mean that the number of animal experiments and the economic costs of developing new chemicals could be reduced. The ability to rapidly pre-screen large and diverse datasets can therefore support the development of new and safer chemicals and help to find replacements for currently used toxic substances. We therefore believe that AI-based methods will help to reduce the negative impact of chemical pollution on humans and ecosystem services," says Erik Kristiansson.

Source: www.chalmers.se