Session I-9: Advances in Three Rs alternatives for reproductive and developmental toxicity Session I-9: Oral presentations

the past decade has seen an enormous increase in the availability of high-information data streams (e.g., genomics) and computational power. these advances, coupled with a strong foundation of toxicological knowledge, have the potential to support remarkable advances in prediction of human risk. this talk will explore a tiered approach to the assessment of the toxicity of chemicals that combines cheminformatics, high-throughput and high-information content data streams, pharmacokinetic considerations, computational modeling and simulation, and relevant in vitro models. Cheminformatic databases are available that contain summaries of all toxicology studies conducted, including over 11,000 entries for reproductive and developmental toxicity. these data can be searched by chemical substructure using expert-based rules to identify analogs of new chemicals. the data can be used to develop a hypothesis about the toxicity of a new chemical, which can often be tested by targeted in vitro studies. High-throughput data streams such as toxCast, along with toxicogenomics, can be used to refine and test hypotheses. A better appreciation of pharmacokinetics and its use in setting appropriate exposure levels in in vitro assays will move beyond simple predictors of hazard to predictors of quantitative risk. Computational simulations are still in relative infancy but have the potential to conduct thousands of virtual experiments, a feature that will be indispensable in identifying the biological pathways that are toxicologically relevant. employing these tools in a tiered fashion will make it possible to evaluate the toxicity potential of many more chemicals than is currently possible, and in a manner that may be more human-relevant. I-9-657 Predicting reproductive and developmental toxicity: A tiered strategy