- Acute toxicity
- Ames test
- animal testing
- Bioconcentration factor
- chemical biological interactions
- chemical hazard
- chemical structure
- Environmental fate
- environmental health perspectives
- EU Chemicals Policy
- Hazard assessment
- high production volume
- In silico approaches
- organic chemicals
- prediction systems
- product development
- Regulation of chemicals
- Risk assessment
- Simplified mathematical representations
- white paper
TR 089 : (Q)SARs: Evaluation of the commercially available software for human health and environmental endpoints with respect to chemical management applications | September 2003
With the marked escalation in demand for information on the potential effects of chemicals on health and the environment, there is a renewed interest in the development and use of (Q)SARs ((quantitative) structure activity relationships) to meet these demands with greater speed, and with less resources. Of particular importance being the potential for such models to minimise the need for animal testing.
Most recently the proposal for a new chemical policy in the European Union (EU (EC, 2001) represent a major challenge for industry, and indeed for all stakeholders, in the numbers of chemicals to be evaluated for their potential to affect adversely human health and the environment.
This report evaluates the current status of the commercially-available (Q)SAR approaches for human health and environmental endpoints, in the context of their applicability to product development and regulatory decision making, such as in hazard assessment (classification) or risk assessment.
(Q)SARs are simplified (mathematical) representations of complex chemical-biological interactions. They can be divided into two major types, QSARs and SARs. QSARs are all quantitative models yielding a continuous or categorical result. The most common techniques for developing QSARs are regression analysis, neural nets and classification methods. Examples of regression analysis include ordinary least squares and partial least squares. Examples of classification methods are discriminant analysis, decision trees and distance based methods of similarity analysis. SARs are qualitative relationships in the form of structural alerts that incorporate molecular substructures or fragments related to the presence or absence of activity.
(Q)SAR predictions are potentially more uncertain than the underlying test data. This imposes limitations on the acceptable use of (Q)SAR in chemical management and decision-making. Approaches to determine the acceptability of (Q)SAR predictions have been developed in the past, but because of their breadth and generality they have not been widely applied or respected by either (Q)SAR users or developers. As a consequence, decision making on the basis of existing models must be done with care and is subject to expert opinion as there is currently no framework for QSAR use and therefore a lack of confidence in these predictions.
Industry is being encouraged to lead the initiative to build more widespread confidence in the appropriate use of (Q)SARs in chemicals management and, towards this goal, to develop jointly with other stakeholders a framework to govern such approaches. This framework should address the acceptability of both negative and positive hazard predictions.
This report reviews the currently-available, most preferred (Q)SAR-based predictive software against acceptability criteria that were developed during the ICCA workshop on regulatory acceptance of QSARs for both human health and environmental endpoints (Cefic, 2002). Only endpoints for which models with sufficiently large databases are available are considered. Difficulties encountered with their modelling are discussed.