Computational Toxicology

Abstract

This chapter briefly explores the principles and application of predictive computational approaches (cheminformatics) to the field of toxicology. In recent years, regulatory and consumer pressures have promoted the use of non‐animal alternatives (including cheminformatics approaches) to assess hazards associated with novel and existing chemicals. In general, these in silico approaches enable us to do two important things: ( 1 ) Rapidly assess likely molecular mechanisms of toxicity; and ( 2 ) Suggest mechanistic hypotheses for experimental validation or refutation. To this end, we provide an overview of current two‐dimensional and three‐dimensional computational approaches including mechanistic machine learning, structural scaffolding, docking, and molecular dynamics simulations. Together with techniques such as inverse docking and pharmacophore/toxicophore mapping, it is possible to use these techniques to align compounds with their potential macromolecular targets, including off‐targets of toxicity. While the approaches discussed in this chapter could potentially be of value when assessing hazards associated with biological xenobiotics and protein–DNA interactions, our focus is on small‐molecule toxicants and their in vivo [protein] targets.