RESUMO
ABSTRACT Hazard identification and health risk assessment traditionally rely on results of experimental testing in laboratory animals. It is a lengthy and expensive process, which at the end still involves large uncertainty because the sensitivity of animals is unequal to the sensitivity of humans. The Agency for Toxic Substances and Disease Registry (ATSDR) Computational Toxicology and Method Development Laboratory develops and applies advanced computational models that augment the traditional toxicological approach with multilevel cross-extrapolation techniques. On the one hand, these techniques help to reduce the uncertainty associated with experimental testing, and on the other, they encompass yet untested chemicals, which otherwise would be left out of public health assessment. Computational models also improve understanding of the mode of action of toxic agents, and fundamental mechanisms by which they may cause injury to the people. The improved knowledge is incorporated in scientific health guidance documents of the Agency, including the Toxicological Profiles, which are used as the basis for scientifically defensible public health assessments.
