The computational prediction of toxicological effects in regulatory contexts : current use and future potential of (Q)SAR tools

National audienceQuantitative and qualitative models describing the relationship between chemical structure and biological activities can be used to predict toxicological effects of chemicals and they are referred to as (Q)SAR models. Since the pioneering work of Corwin Hansch in the 60's, the development and utilization of structure-activity relationships have become increasingly more important over the past years for industrial and regulatory applications. The implementation of new European chemical safety policies (REACH, 7(th) amendment of the EU cosmetic Directive) is one of the incentive of this trend whose success is dependent upon the implementation of in silico methods among fully integrated strategies. This review briefly summarizes the current utilization and future potential of such modeling approaches for predicting chemical-induced human health hazard