In silico risk assessment for skin sensitization using artificial neural network analysis.

摘要: The sensitizing potential of chemicals is usually identified and characterized using in vivo methods such as the murine local lymph node assay (LLNA). Due to regulatory constraints ethical concerns, alternatives animal testing are needed predict skin sensitization chemicals. For this purpose, an integrated evaluation system employing multiple vitro silico parameters that reflect different aspects process seems promising. We previously reported LLNA thresholds could be well predicted by artificial neural network (ANN) model, designated iSENS ver. 2 (integrating tests version 2), analyze data obtained from focused on sensitization. Here, we examined whether ANN silico-calculated descriptors three-dimensional structures a good correlation between values. Furthermore, combining results (iSENS 2) models reduced number for which potency category was under-estimated. In conclusion, model shown have useful predictive performance. Further, our indicate combination with represents promising approach risk assessment