Global versus local QSAR models for predicting ionic liquids toxicity against IPC-81 leukemia rat cell line: The predictive ability

Abstract

Abstract The unlimited possibility of modifying the structure of the cation and anion of a given ionic liquid (IL) delivers a countless number of potential derivatives having different toxicological activities. The experimental measurements for such huge set of compounds are impossible due to time and costs limitations. In the absence of experimental data, computational methods such as quantitative structure–activity relationship (QSAR) method can be applied to characterize those species for which experimental data is not available. However, two different ways of QSAR modeling could be applied: local modeling - dedicating to one specific group and more complex global models. The aim of this study was to investigate how local/global QSAR modeling works for the ionic liquids. Which of these two approaches is more suitable in ILs' modeling? We have developed six local models and a global one for predicting the toxicity against the leukemia rat cell line IPC-81. Based on the obtained results, we conclude that local models are useful to solve specific mechanistic problems among one group of ILs, but in case of predicting the toxicity for a new designed compound the global model, whenever fulfills all quality recommendations, is recommended.