AI-PolyRisk - Artificial Intelligence to Prioritize and Identify Polymers as New Emerging Risk Chemicals

Chemicals in products and materials provide large societal and health benefits, however some chemicals are ubiquitous in the environment and some have shown to cause adverse effects in biota including humans.

Sustainable use of chemicals and efficient regulatory and policy actions requires expertise and tools for early identification of potentially hazardous chemicals.

This project aims at developing tools to characterize and identify polymers as new and/or emerging risk chemicals (NERCs).

Polymers, often overlooked in risk assessment of chemicals, will be studied combining experimental environmental fate and hazard studies with AI methodologies for smart text mining of big data and machine learning based modelling.

In particular we will challenge major limitations in available methodologies, including time-consuming data curation, inappropriate use of available data including so-called grey literature, and lack of modelling and experimental approaches and expertise to assess environmental risks of polymers.

The project will provide scientific impact on the development of systematic approaches to identify polymeric NERCs using per- and polyfluoroalkyl substances (PFAS) as case class of chemicals to build up new machine learning, text mining and experimental approaches.

Results are needed for national and international regulators and policymakers to address EU policy under the Green Deal and the Chemicals Strategy aiming for zero pollution and a toxic free environment.