|
(M/EH092) Multicriteria decision-making and QSAR approaches for the screening of Long Range Transport potential of organic pollutants. Gramatica, Paola1, Consolaro, Federica1, Pozzi, Stefano2, Pavan, Manuela1, 1 2 ABSTRACT- Several pollutants of concern, including pesticides, PAHs, PCBs etc. have a Long Range Transport potential due to the combination of their persistence in the environment and their inherent tendency towards "mobility". Different physico-chemical properties are of importance in determining LRT (for instance: volatility, solubility in water, different sorption coefficients, etc.) and the half-life in different compartments is commonly used as a persistence indicator. Finding the best combination of chemical properties minimizing LRT is a multicriteria problem that can be approached by MultiCriteria Decision-Making (MCDM) techniques. Different functions (such as for instance, desirability and utility) are applied to the properties most relevant in determining the Long Range Transport: this allows a ranking of the studied chemicals according to their LRT potential. The QSPR (Quantitative Structure-Property Relationships) approach is applied in two steps: 1) to fill the gap in the experimental data of the studied properties; 2) to model the final scores of the MCDM functions. Different kinds of theoretical molecular descriptors (easily calculable by a free-downloadable web-software) have been used to obtain OLS regression models with predictive power (validated by Q2LOO and Q2LMO); the reliability of predicted data in relation to the chemical domain of the training set is always checked by the leverage approach. The application of QSPR models obtained from MCDM scores would allow a fast screening of both existing and potential chemicals, based simply on the knowledge of their molecular structure, to assess the inherent tendency of such chemicals towards LRT. Key words: POPs, persistence, molecular descriptors, desirability functions |
