Computational toxicology, friend or foe?

There is increasing public pressure to reduce animal testing and yet maintain public safety from exposure to chemicals either in the environment we live in, the food that we eat or the drugs that we take to treat illnesses. Computational approaches offer the attraction of being both fast and cheap to run being able to process thousands of chemical structures in a few minutes. As a result these approaches have seen an increase in interest and effort over the last decade most notably in the pharmaceutical industry where costs for new drug development is soaring and the failure rate for safety reasons is high. Many applications and approaches have been published covering a wide variety of different human and environmental health issues. As with all new technology, there is a tendency for these approaches to be hyped up and claims of reliability and performance may be exaggerated. So just how good are these computational methods? This review is intended to provide an overview of the state of the art in computational toxicology and to illustrate where some of the limitations of these approaches exist so that these valuable tools are applied and interpreted correctly.