LAUSR

Quantitative structure‐toxicity relationship (QSTR) models were built for two in vitro endpoints: cytotoxicity and enzymatic activity of diverse chemicals to goldfish (Crassius auratus) scale tissue (GFS) and topminnow (Poeciliopsis lucida) hepatoma cell line (PLHC‐1), respectively. The data sets were based on experimental cytotoxicity measured with uptake of 3‐amino‐7‐dimethylamino‐2‐methylphenazine hydrochloride dye (Neutral Red assay) representing lysosomal damage and enzymatic activity measured with Ethoxyresorufin‐O‐deethylase (EROD) induction potency. The descriptors were calculated with DRAGON 6 and SPARTAN 10 software packages. Descriptor selection was made by ‘All Subset’ and Genetic Algorithm‐based features implemented in QSARINS software. The proposed QSTR models validated both internally and externally. Additionally, the QSTR models generated for cytotoxicity and EROD induction potency were used to predict the relevant endpoint values for external set chemicals with structural coverage of 95.0 % and 92.1 %, respectively. A strong correlation of experimental in vivo fish lethality data with predicted in vitro cytotoxicity and EROD induction potency values for external set chemicals was found. It was concluded that the proposed QSTR models might be useful to provide an initial screening and prioritization for these diverse chemicals. Also, regarding the strong correlations between predicted in vitro and experimental in vivo data, the use of QSTR predictions as an alternative to the acute fish toxicity assessment can be claimed.