Image: Kristine Grønning Kongsbak

Computers predict effects of chemicals on human health

Food, fish and agriculture Food safety Computer calculations Biological systems

Computer models can group chemicals according to their health effects and can contribute to calculating and predicting cocktail effects on human health. These are some of the findings from a PhD project from the National Food Institute, Technical University of Denmark.

Every day people are exposed to several chemicals through foods, cosmetics, indoor air and pharmaceuticals. Chemical exposure is thought to play an important role in the development of several adverse health effects in humans. However, testing all potential combinations of the vast number of available chemicals that people are exposed to would be time-consuming and impractical.

In her PhD project which is the result of a joint effort from the National Food Institute and DTU Systems Biology Kristine Kongsbak has studied how existing computer models from systems biology can be used in toxicological research to predict potential health effects of various chemicals - both in isolation and when they occur in cocktails.

Systems biology is an approach to biology which focuses on the complex interactions in biological systems.

New knowledge can target studies

The computer models can establish to which proteins specific chemicals bind and elucidate the pathways that are involved in the effects of the chemicals, before finally creating a link to the human diseases the chemicals may cause. This can be done by focusing exclusively on human effects, which avoids problems of having to predict human effects from animal studies.

The hypotheses which are generated can inspire the design and planning of new studies of chemicals’ adverse effects. As such the use of computer models is not intended to replace classical toxicology, but rather to be a useful supplement. 

Grouping chemicals according to their effects

The PhD project has also shown how computer models can group chemicals according to the adverse effects that are predicted to occur in humans. The models use data from previous experiments, which are available through public databases.

The more and better data is available, the better the predictions of the effects according to which the chemicals are grouped will become. As such the method can be expected to become even more valuable as more data will become available in the future.

Read more

Read Kristine Kongsbak’s PhD thesis: Integrative Systems Biology Applied to Toxicology (pdf).

The PhD was financed by the Ministry of Food, Fisheries and Agriculture in a larger project on the management of chemical cocktail effects, Cocktail. The aim of the project along with the main results have been described in more detail in the booklet: Cocktail: Danmarks største forskningsprojekt om cocktaileffekter i fødevarer (pdf – available in Danish only).

Also read National Food Institute's press release from 19 March 2015: New knowledge strengthens risk assessment of chemical cocktails in food.