You are here:
CURRENT PRACTICES IN QSAR DEVELOPMENT AND APPLICATIONS
Citation:
Richard, A M. CURRENT PRACTICES IN QSAR DEVELOPMENT AND APPLICATIONS. Presented at Toxicology and Risk Assessment Approaches for the 21st Century, Kings Island Resort, OH, April 10-12, 2000.
Description:
Current Practices in QSAR Development and Applications
Although it is commonly assumed that the structure and properties of a single chemical determines its activity in a particular biological system, it is only through study of how biological activity varies with changes in chemical structure across a series of chemicals that usefull generalizations and associations are developed. Quantitative structure-activity relationships (QSARs) provide formallinl(ages of chemical information, implicit in molecular structure, to the quantitative and qualitative expression of this information in biological systems. The multitude of ways in which these linkages can be developed, through statistical analysis on a global scale (universe of chemicals) or mechanism-based analysis on a local (class of chemicals) scale, what they can potentially reveal (general associations of structural features, key reactivity elements, insight into mechanisms), and how this information can be used (to generate or refine mechanistic hypotheses, to predict the potential activity of new chemicals, to support risk assessment) span applications across the spectrum of toxicological inquiry. A brief overview of concepts governing the construction of useful QSARs will be provided, emphasizing the importance of both statistical significance and chemical and biological plausibility. In addition, some current examples of QSAR application will be discussed, ranging from evaluation of a "general carcingenicity screening" model recently developed by the FDA using the CASE system, to consideration of the increasing role of computational chemistry in helping to better define the chemical reactivity component of the QSAR paradigm.
This abstract is not neccessarily reflect US EPA policy.