Computational Tools for the Prediction and Classification of Estrogenic Compounds

EPA Grant Number: R826133
Title: Computational Tools for the Prediction and Classification of Estrogenic Compounds
Investigators: Welsh, William J.
Institution: University of Missouri - St Louis
EPA Project Officer: Deener, Kacee
Project Period: January 1, 1998 through December 31, 2000
Project Amount: $433,758
RFA: Endocrine Disruptors (1997) RFA Text |  Recipients Lists
Research Category: Economics and Decision Sciences , Endocrine Disruptors , Health , Safer Chemicals


The primary objective of this research is to develop an integrated array of computational tools for use in research and regulation of potential endocrine disrupting compounds (EDCs). My research team has joined with scientists at the National Center for Toxicological Research (NCTR) to construct and validate models based on Quantitative Structure-Activity Relationship (QSAR) approaches for the classification, identification, and prediction of potential EDCs. We will also employ a rules-based expert system capable of identifying possible metabolites of parent compounds which in turn can be submitted to our QSAR models for evaluation. We will also explore correlations between estrogenic activities predicted by our QSAR models for one species and experimental activities for another species (inter-species extrapolation). Establishing quantitative relationships between species (e.g., mouse, rat, calf, human) will greatly benefit efforts in risk assessment and will enable us to pool biological assay data from different species and, thereby, to expand the database of chemicals for building our QSAR models.


Our proposed testing scheme combines two QSAR paradigms (classical descriptor-based QSAR and 3D-QSAR (i.e., CoMFA)), three programs for generating >400 molecular descriptors including many parameters (e.g., log P, Gsolvation in aqueous and organic media) necessary for modeling in vivo activity, both linear (Partial Least Squares) and non-linear (Artificial Neural Networks) regression models, and Genetic Algorithms (GAs) to select an optimal set of molecular descriptors for modeling and predicting estrogenicity. We will extend and organize our QSAR models to predict estrogenic activity in vitro across the hierarchy of molecular-level biological complexity, beginning with ligand-estrogen receptor binding and progressing toward gene regulation and transcription.

Expected Results:

These QSAR models are designed to help in prioritizing the anticipated large number of chemicals for extensive in vitro and in vivo testing based on their predicted estrogenicity.

Publications and Presentations:

Publications have been submitted on this project: View all 68 publications for this project

Journal Articles:

Journal Articles have been submitted on this project: View all 15 journal articles for this project

Supplemental Keywords:

endocrine disruptors (endocrine disrupting compounds), risk assessment, Quantitative Structure-Activity Relationships (QSARs), estrogens., RFA, Health, Scientific Discipline, Health Risk Assessment, Environmental Chemistry, Endocrine Disruptors - Environmental Exposure & Risk, Risk Assessments, Analytical Chemistry, endocrine disruptors, Biochemistry, Children's Health, Biology, Endocrine Disruptors - Human Health, adverse outcomes, risk assessment, expert systems, metabolites, computational tool, Quantitative Structure-Activity Relationship, human exposure, animal models, developmental processes, estrogen response

Progress and Final Reports:

Final Report