【编者按】网学网应用化学频道为大家收集整理了“环境污染物毒性的QSAR研究“提供大家参考，希望对大家有所帮助!
论文编号:HG092 论文字数:11513,页数:24
摘 要：卤代芳烃类有机污染物在环境中分布广泛，对生态系统有极大的危害。在评价卤代芳烃类化合物在环境中的传播时，它的理化性质对于卤代多环芳烃类化合物的环境暴露评价、生态风险和人类健康评价，具有重要意义。定量结构-活性相关(即构效关系，简称QSAR)是建立有机物的活性（在环境化学中指化合物的毒性）与表征其结构特征的理化参数之间的相关性方程，通过测量或计算有机物的理化参数，可大体估算出有机物对生物的毒性。MATLAB中的神经网络工具箱是进行神经网络系统分析与设计的有力工具。RBF神经网络以其计算量小，学习速度快，不易陷入局部极小等诸多优点为系统辨识与建模提供了一种有效的手段。将二者结合起来，解决了卤代烷烃烯烃化合物的建模问题。建立多氯代烃类化合物的QSAR模型，可以预测氯代烃类化合物的性质，为研究多卤代烃类化合物的环境行为提供可借鉴的方法和手段，为此类化合物的生态风险评价提供数据基础，为进一步探讨分子的结构与性质之间的关系提供理论参考。
关键词：卤代烃类化合物；定量结构活性关系；RBF神经网络
QSAR for Acute Toxicity of Environmental Pollutants
Abstract: The halogenated hydrocarbons distribute widely in the environment,and have huge harm to the environment and biogeocenose .Physicochemical properties of an organic chenmical compound play an important role in determining its distribution and fate in the environment.QSAR methods,are the most promising and successful tools to provide rapid and useful meaning for predicting the biological toxicities of organic compounds by use various kinds of molecular descriptors.The goals of QSAR is to develop models on a training set of compounds,these models will then allow for the prediction of the biological activity of related chemicals. This kind of study can not only develop a method for the prediction of the property of compounds,thus gain some insight into instrutural factors affecting molecular propertise.Neural network system toolbox is a powerful tool for analysis and design in MATLAB. RBF neural network has many advantages:small amount of calculation,learn fast, and so on. By conbining the two together,the modelling of halogenated hydrocarbons is resolved and achieved satisfactory results.QSAR will give a useful method to study its distribution and fate in the environment,at the same time,QSAR afford the date of physicochemical properties of an organic chemicl compound.
Keywords:halogenated hydrocarbons; Quantitative Structure-activity Relationship (QSAR); RBF neural network
目 录
中文摘要 I
英文摘要 II
目录 III
1. 绪论 1
1.1 卤代化合物的毒性 1
1.1.1 卤代化合物的环境影响分析 1
1.1.2 取代基对化合物毒性的影响 2
1.2 定量结构活性关系简介 4
2. 实验部分 13
2.1 多元线性逐步回归法构建数学模型 13
2.2 神经网络法构建模型 16
3. 结果与讨论 18
4．总结与展望 19
致谢 20
参考文献 21
