Predictive Models and Computational Toxicology

摘要: Understanding the potential health risks posed by environmental chemicals is a significant challenge elevated large number of diverse with generally uncharacterized exposures, mechanisms, and toxicities. The ToxCast computational toxicology research program was launched EPA in 2007 part federal Tox21 consortium to develop cost-effective approach for efficiently prioritizing toxicity testing thousands application this information assessing human toxicology. addresses problem through an integrated workflow using high-throughput screening (HTS) chemical libraries across more than 650 vitro assays including biochemical assays, cells cell lines, alternative models such as mouse embryonic stem zebrafish embryo development. initial phase profiled library 309 chemicals, mostly pesticidal actives having rich vivo data from guideline studies that include chronic/cancer bioassays mice rats, multigenerational reproductive prenatal developmental endpoints rats rabbits. first used build aim determine how well animal effects can be predicted solely data. Phase I now complete both (ToxCast) anchoring database (ToxRefDB) have been made available public (http://actor.epa.gov/). As II underway, purpose chapter review progress date predictive modeling, specific examples on rabbits lessons learned during I.