The computational prediction of toxicity

Recent developments in the prediction of toxicity from chemical structure h ave been reviewed. Attention has been drawn to some of the problems that ca n be encountered in the area of predictive toxicology, including the need f or a multi-disciplinary approach and the need to address mechanisms of acti on. Progress has been hampered by the sparseness of good quality toxicologi cal data. Perhaps too much effort has been devoted to exploring new statist ical methods rather than to the creation of data sets for hitherto uninvest igated toxicological endpoints and/or classes of chemicals.