Introducing Membrane Transport Energy into the Design of Sustainable Chemicals against Cytotoxicity

摘要

Adverse outcomes associated with chemical use and spill have raised concerns over its sustainability. The traditional toxicity testing, using laboratory animals as the downstream safety check, is becoming less viable due to the economical and ethical liabilities. Green chemistry proposed an alternative strategy to attain chemical sustainability: designing chemicals to maximize their intrinsic sustainability and thus minimize their hazardous risk. Following decades of progress, there is still a need to develop new metrics to quantify the level of chemical sustainability. In this report, we developed a new double functional tool capable of estimating the sustainability probability of a chemical and designing new chemicals to meet a desired sustainability probability. This tool was built upon the Naive Bayesian algorithm with the design variables stemming from three sources. Molecular softness and polarizability were derived from density functional theory (DFT), and membrane transport free energy was computed using our in-house developed finite element algorithm. Model validation against the cytotoxicity measured in the U.S. EPA Toxicity ForeCaster (ToxCast) database (tested up to 100 mu M) yielded a score of 0.82 for the area under the curve (AUC) of the receiver operating characteristic (ROC). On the basis of this model, we constructed the assessment tool with the dual capabilities of prediction and design.