论文部分内容阅读
定量构效关系(QSAR)模型在研究饮用水消毒副产物(DBPs)中氯原子数量与毒性的关系方面有一定进展,然而,能够反映DBPs毒理性的分子描述符和氯原子数量之间的相关性却从未被系统地和定量的研究。分子描述符的计算是QSAR研究的基础,其中分子轨道能,如最低空轨道能量(ELUMO)是应用最广的量化参数之一。基于QSAR方法建模筛选氯代烯烃同系物的相关量子化学参数,检验最高占据轨道能级(EHOMO)、碳原子数(NC)、氯原子数(NCl)对ELUMO的贡献作用。并通过统计学方法进行模型开发和内外交叉法验证发现,NCl和NC是预测氯代烯烃分子特性较重要的两个化学参数,即较稳健的模型为ELUMO=-0.0205NCl+0.0059NC+0.1641(n=11,R2=0.9956,F=102.8569,RMSE=0.0019)。此外,为了提高模型的准确性,使用蒙特卡罗法对模型进行不确定性分析,来进一步评估这些参数的误差源传输到ELUMO计算时,对ELUMO的影响。
Quantitative Structure-Activity Relationship (QSAR) models have made some progress in studying the relationship between the number of chlorine atoms in drinking water disinfection by-products (DBPs) and toxicity. However, the relationship between the molecular descriptors and the number of chlorine atoms that can reflect the toxicity of DBPs Relevance has never been systematically and quantitatively studied. The calculation of molecular descriptors is the basis for the QSAR study, in which molecular orbital energies, such as the lowest empty orbital energy (ELUMO), are among the most widely used quantitative parameters. Based on QSAR method, the related quantum chemical parameters of chlorinated olefin homologues were screened and the contribution of highest occupied orbital energy (EHOMO), number of carbon atoms (NC) and number of chlorine atoms (NCl) to ELUMO was tested. And through the statistical methods of model development and internal and external cross validation, NCl and NC are two important chemical parameters for predicting molecular properties of chloroolefins. The more robust models are ELUMO = -0.0205NCl + 0.0059NC + 0.1641 ( n = 11, R2 = 0.9956, F = 102.8569, RMSE = 0.0019). In addition, in order to improve the accuracy of the model, the Monte Carlo method is used to analyze the uncertainty of the model to further evaluate the impact on the ELUMO when the error sources of these parameters are transmitted to the ELUMO calculation.