QSAR models for the screening, prediction and refinement of PBT Properties of Contaminants of Emerging Concern.

The prompt identification of the adverse effects of Contaminant of Emerging Concern (CEC) is fundamental to ensure high protection level for human health and the environment. Persistent, Bioaccumulative and Toxic (PBT) compounds are chemicals of high concern and should be readily identified. The aim of this thesis is to propose an approach based on Quantitative Structure Activity Relationship (QSAR) models for the evaluation of the intrinsic environmental hazard of CECs. First, a screening of the potential PBT behavior of pharmaceuticals is performed by consensus approach. Results demonstrate a high agreement (i.e.86%) between the different QSAR models. Then, QSARs are developed to estimate acute toxicity of pharmaceuticals in aquatic species. All models have good fitting (R2>0.75) and predictivity (Q2EXT>0.68). An Aquatic Toxicity Index is proposed and modelled. Moreover, interspecies correlation models are also developed. Finally, QSARs for the prediction of whole-body human biotransformation Half-Lives are developed for organic chemicals. Predictions for the biotransformation potential are integrated in a mechanistic mass-balance multimedia environmental fate food-web model to estimate the Biomagnification Factor (BMF) in human in a tiered approach. The introduction of biotransformation strongly affects the calculation of BMF and the elimination processes related to biotransformation are predominant in the overall bioaccumulation.