Predicting the skin sensitizing potential of pesticides using Pred-skin 3.0–A web-based prediction tool

Abstract

Pesticide usage is increasing due to growing needs of agriculture and horticulture. Occupational dermal exposure to pesticides at an acute or chronic low-level could result in contact dermatitis and various skin cancers. Hence, detailed understanding about the Adverse Outcome Pathways (AOP) or Chemical Sensitization Pathway (CSP) behind pesticides belonging to various categories has to be investigated. Animal models of skin sensitization testing at times either over or under predict the human responses due to species-to-species variability. This necessitates the need for prediction tools for skin sensitizing potential of various chemicals. Pred-skin 3.0, is a consensus Naïve Bayes model-based prediction tool which utilizes various human, LLNA, and non-animal data to predict skin sensitization. Although, this tool was never used for predicting skin sensitizing potential of pesticides. Henceforth, the current study aims to test the applicability of this prediction tool in predicting skin sensitizing potential of 96 pesticides belonging to three Major classes. The Bayesian outcome of Pred Skin prediction tool provided a good concordance of 72.72 % with the existing animal skin sensitizing data as well as 63.46 % with the non-sensitizer data.