Chen Xingmei, Dang Limin, Yang Hai, Huang Xianwei, Yu Xinliang
Hunan Provincial Key Laboratory of Environmental Catalysis & Waste Regeneration, College of Materials and Chemical Engineering, Hunan Institute of Engineering Xiangtan Hunan 411104 China
RSC Adv. 2020 Oct 1;10(59):36174-36180. doi: 10.1039/d0ra05906d. eCollection 2020 Sep 28.
Predicting the acute toxicity of a large dataset of diverse chemicals against fathead minnows () is challenging. In this paper, 963 organic compounds with acute toxicity towards fathead minnows were split into a training set (482 compounds) and a test set (481 compounds) with an approximate ratio of 1 : 1. Only six molecular descriptors were used to establish the quantitative structure-activity/toxicity relationship (QSAR/QSTR) model for 96 hour LC through a support vector machine (SVM) along with genetic algorithm. The optimal SVM model ( = 0.756) was verified using both internal (leave-one-out cross-validation) and external validations. The validation results ( = 0.699 and = 0.744) were satisfactory in predicting acute toxicity in fathead minnows compared with other models reported in the literature, although our SVM model has only six molecular descriptors and a large data set for the test set consisting of 481 compounds.
预测大量不同化学物质对黑头呆鱼（）的急性毒性具有挑战性。在本文中，将963种对黑头呆鱼具有急性毒性的有机化合物按近似1∶1的比例分为训练集（482种化合物）和测试集（481种化合物）。仅使用六个分子描述符，通过支持向量机（SVM）结合遗传算法，建立了96小时半数致死浓度（LC）的定量构效/毒性关系（QSAR/QSTR）模型。使用内部验证（留一法交叉验证）和外部验证对最优SVM模型（ = 0.756）进行了验证。尽管我们的SVM模型仅有六个分子描述符且测试集有一个由481种化合物组成的大数据集，但与文献中报道的其他模型相比，验证结果（ = 0.699和 = 0.744）在预测黑头呆鱼的急性毒性方面令人满意。
