Multi-criteria decision-making techniques associated with (Q)SAR risk assessment for ranking surface water...

Contaminants of emerging concern (CECs) have been a focus of study for years, with investigations revealing the contamination of different environmental matrices (surface water, soil, air, and sediment) by diverse classes of microcontaminants. Understanding the contamination profiles requires identification and risk assessment of the microcontaminants. In the present work, analysis was made of the presence of 3250 compounds in 27 samples from the Conceição River (Rio Grande do Sul State, Brazil), using an SPE-LC-QTOF MS method. In total, 150 microcontaminants (confirmed and suspected) of different classes, especially pesticides and pharmaceuticals, were identified by an initial qualitative analysis. Subsequently, in silico predictions of eight endpoints, using quantitative structure-activity relationship ((Q)SAR) models, were employed to determine the risk of each previously screened microcontaminant. This large amount of (Q)SAR data, frequently with conflicting information in relation to the responses of the different endpoints, makes it difficult to define which microcontaminants should be prioritized for analysis. Therefore, in order to rank the identified microcontaminants by risk assessment, two multi-criteria decision-making (MCDM) ranking techniques (ToxPi and TOPSIS), associated with a weighting method, were performed to establish the order of priority for further quantitative analysis of the most hazardous microcontaminants. The two rankings were statistically similar, especially for the 20 highest priority microcontaminants. Nonetheless, sensitivity tests carried out for the ToxPi and TOPSIS outputs showed higher performance robustness of TOPSIS, compared to ToxPi. This is the first time that such an approach (screening/(Q)SAR/MCDM methods) has been performed in the context of microcontaminant environmental risk evaluation and demonstrated to be an available strategy to help rank the most concern microcontaminants identified in aqueous environment samples.