Establishment of an evaluation method of a chemical-susceptibility for wildlife using counter-QSAR and machine learning

Certain environmental chemicals are harmful to lifeforms. Determining the extent to which they are damaging to animals is challenging as it is ethically reprehensible to expose animals to lethal doses of chemicals in the lab and, besides, not every species of animal can be housed in a laboratory. So, how can experiments involving animals in their natural habitat, and without causing harm, be conducted? A team of researchers at the Laboratory of Toxicology in the School of Veterinary Medicine at Kitasato University in Japan, which includes Junior Associate Professor Kazuki Takeda, is working to develop a way to determine the susceptibility of animal species to given environmental chemicals. Takeda is applying his background in environmental toxicology to the research. The focus for the team is on wild animals and using non-invasive methods. As there is such great interspecies variability in chemical sensitivity, it is difficult to extrapolate toxicity test results obtained from laboratory animals to wild animals. Therefore, the researchers are using genetic information and protein-chemical binding simulations to evaluate interspecies differences in chemical sensitivity. The goal is to develop a model that can predict chemical sensitivity differences between species using simulation methods and machine learning using DNA information obtained from hair.