Development of a classification model for the antigenotoxic activity of flavonoids

Abstract Genotoxic agents are capable of causing damage to genetic material and the cumulative DNA damage causes mutations, involved in the development of various pathological conditions, including cancer. Antigenotoxic agents possess the potential to counteract these detrimental cellular modifications and may aid in preventing, delaying, or decreasing the severity of these pathological conditions. An important class of natural products for which promising antigenotoxic activities have already been shown, are the flavonoids. In this research, we investigated the quantitative structure-activity relationship (QSAR) of flavonoids and their antigenotoxic activity against benzo[a]pyrene (B[a]P) and its mutagenic metabolite B[a]P-7,8-diol-9,10-epoxide-2. Random Forest classification models were developed, which could be useful as a preliminary in silico evaluation tool, before performing in vitro or in vivo experiments. The descriptors G2S and R8s. were the most significant for predicting the antigenotoxic potential.