1st International Congress on Environmental Modelling and Software - Lugano, Switzerland - June 2002
Keywords
pops, lrt, qsar, classification, multicriteria
Start Date
1-7-2002 12:00 AM
Abstract
The Long Range Transport (LRT) potential of chemicals is due to the combination of theirpersistence in the environment and their inherent tendency towards mobility, and is an undesirableproperty of POPs (Persistent Organic Pollutants). Finding the best combination of chemical properties tominimize LRT is a multicriteria problem that can be approached by MultiCriteria Decision-Making(MCDM) techniques. Utility functions have been applied to two proposed indexes, the “globalpersistence index” and the “mobility index”, allowing a ranking of the studied chemicals according totheir LRT potential. The “global persistence index” was obtained by linear combination, by PrincipalComponent Analysis of the half-life data in various environmental compartments. Half-life data arecommonly used as persistence indicators, but the availability of such data is limited to only a few organiccompounds, thus QSAR (Quantitative Structure-Activity Relationships) models were used to predict suchdata starting from molecular structure information. Validated OLS regression models were realised usingdifferent theoretical molecular descriptors able to predict half-life values; all the regression models werestrongly validated for their prediction power by the leave-one-out and leave-more-out procedure (Q2=70-90%). Analogously the “mobility index” was ascertained by Principal Component Analysis of differentphysico-chemical properties relevant to the determining of LRT, for instance, volatility, water solubilityand different sorption coefficients. The final application of QSAR regression models and classificationmodels (CART) on the obtained MCDM scores allows a fast screening and ranking of existing chemicalsfor their inherent tendency towards LRT. This approach could be usefully applied also to new chemicals,even those not yet synthesized, as it is based simply on the knowledge of the molecular structure.
Modelling Of POP Environmental Persistence And Long Range Transport By QSAR And Chemometric Approaches
The Long Range Transport (LRT) potential of chemicals is due to the combination of theirpersistence in the environment and their inherent tendency towards mobility, and is an undesirableproperty of POPs (Persistent Organic Pollutants). Finding the best combination of chemical properties tominimize LRT is a multicriteria problem that can be approached by MultiCriteria Decision-Making(MCDM) techniques. Utility functions have been applied to two proposed indexes, the “globalpersistence index” and the “mobility index”, allowing a ranking of the studied chemicals according totheir LRT potential. The “global persistence index” was obtained by linear combination, by PrincipalComponent Analysis of the half-life data in various environmental compartments. Half-life data arecommonly used as persistence indicators, but the availability of such data is limited to only a few organiccompounds, thus QSAR (Quantitative Structure-Activity Relationships) models were used to predict suchdata starting from molecular structure information. Validated OLS regression models were realised usingdifferent theoretical molecular descriptors able to predict half-life values; all the regression models werestrongly validated for their prediction power by the leave-one-out and leave-more-out procedure (Q2=70-90%). Analogously the “mobility index” was ascertained by Principal Component Analysis of differentphysico-chemical properties relevant to the determining of LRT, for instance, volatility, water solubilityand different sorption coefficients. The final application of QSAR regression models and classificationmodels (CART) on the obtained MCDM scores allows a fast screening and ranking of existing chemicalsfor their inherent tendency towards LRT. This approach could be usefully applied also to new chemicals,even those not yet synthesized, as it is based simply on the knowledge of the molecular structure.
