QSAR models are mainly useful in the prediction of data for chemicals without experimental information, those not yet tested or even not yet synthesized. However, QSAR models must be carefully verified for their reliability in the specific context of application to be successfully used and not misused. The OECD principles for QSAR model validation have established crucial points for prediction reliability, mainly model reproducibility, external predictivity and applicability domain checking. Particular attention must be paid to QSAR models’ predictive performance, when the models are applied to the screening of large chemical sets, the specific aim being to focus on the most hazardous to prioritize them for experimental tests. Different approaches for splitting the available experimental data sets and various statistical parameters can be used to verify a model’s external predictivity. These fundamental aspects of QSAR model reliability are commented on, based on several examples of application to various environmental organic pollutants, such as Persistent, Bioaccumulative and Toxic (PBTs) chemicals, Endocrine Disruptors (EDs), flame retardants and polyfluorinated chemicals. Some of these compounds are included in the priority list of Persistent Organic Pollutants (POPs) and/or are among the substances of very high concern (SVHCs), which require authorization in REACH. Therefore methods are needed for an early identification of these pollutants. The QSAR models presented could be particularly useful for screening and prioritisation purposes, also a priori in a green chemistry approach, in the design of new products as safer alternatives to existing dangerous chemicals (“benign by design”).

Predictive QSAR modelling for Screening and Prioritization of Environmental Organic Pollutants

GRAMATICA, PAOLA
2011-01-01

Abstract

QSAR models are mainly useful in the prediction of data for chemicals without experimental information, those not yet tested or even not yet synthesized. However, QSAR models must be carefully verified for their reliability in the specific context of application to be successfully used and not misused. The OECD principles for QSAR model validation have established crucial points for prediction reliability, mainly model reproducibility, external predictivity and applicability domain checking. Particular attention must be paid to QSAR models’ predictive performance, when the models are applied to the screening of large chemical sets, the specific aim being to focus on the most hazardous to prioritize them for experimental tests. Different approaches for splitting the available experimental data sets and various statistical parameters can be used to verify a model’s external predictivity. These fundamental aspects of QSAR model reliability are commented on, based on several examples of application to various environmental organic pollutants, such as Persistent, Bioaccumulative and Toxic (PBTs) chemicals, Endocrine Disruptors (EDs), flame retardants and polyfluorinated chemicals. Some of these compounds are included in the priority list of Persistent Organic Pollutants (POPs) and/or are among the substances of very high concern (SVHCs), which require authorization in REACH. Therefore methods are needed for an early identification of these pollutants. The QSAR models presented could be particularly useful for screening and prioritisation purposes, also a priori in a green chemistry approach, in the design of new products as safer alternatives to existing dangerous chemicals (“benign by design”).
2011
Gramatica, Paola
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11383/1727787
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact