Defining species sensitivities to endocrine disrupting chemicals.

The goal of ecological risk assessment is to evaluate the risks that human-released substances pose to living organisms and ecosystems. It raises a key question for scientists, regulators, and the public: Are current ecological risk assessment methods effective? Unfortunately, the evidence suggests they are not. We are experiencing significant biodiversity loss across all ecosystems due to climate change, habitat destruction, and pollution.
Chemical pollutants are everywhere, found in our oceans, mountains, and various habitats. We have surpassed the safe limit for chemicals and new substances that humanity can manage. There are hundreds of thousands of chemicals in use today, and this number keeps growing, making it challenging to identify which chemicals are most harmful and which species are most at risk. One group of chemicals of concern are those that interfere with the endocrine system because this system is vital for well-being and sexual development.
Despite the daunting task of evaluating the risk of so many chemicals, new computational approaches using the extensive toxicological, chemical, protein, and genomic information offers promising solutions. This project aims to create new methodologies to identify species that are most sensitive to endocrine-disrupting chemicals, as well as identify the chemicals of greatest concern. We will use computational tools to predict how strongly chemicals bind and activate nine types of nuclear receptor families vital for endocrine system function and health that are found in all vertebrates, such as fish, amphibians, reptiles, birds, and mammals.
Initially, we will use computer methods to predict how chemicals bind to these receptors. We will then compare our findings with lab tests that measure how chemicals interact with receptors. This will provide accurate binding strength and gene activity data. We can use this information to improve our predictive models with artificial intelligence techniques, like deep neural networks. Our data will also help identify key protein features that indicate a species' sensitivity based on protein structure. This approach could be extended to other chemical targets, as well as invertebrates, plants, and algae. This can help identify chemicals of concern, whether they are legacy pollutants or emerge during the design of chemical products intended to reduce or eliminate the use of hazardous substances.