Studies On Quantitative Structure-Activity(Toxicity) Relationships For Some Complex Organic Pullutants

Quantitative structure–activity relationship(QSAR) methods, are the most promising and successful tools to provide rapid and useful meanings for predicting the biological activity or toxicities of organic compounds by use of different statistical methods and 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 or toxicities of related chemicals. This kind of study can not only develop a method for the prediction of the property of compounds that have been synthesized but also can identify and describe important structural features of moleculars that are relevant to variations in molecular propertise, thus gain some insight into instrutural factors affecting molecular propertise. Now, QSAR method has been introduced to biology, chemistry and environment science. In the dissertation, we mainly discuced the application of QSAR in the study on the biological activity or toxicities of some complex organic pullutants in the environment, according to molecular descriptors of struture and establish the relationships between structure and activity(toxicity).Compound structures determine their nature and each character is closely related to their structures. Therefore, it is great significant that analysis and study their change rules between molecular structure characterizations and physico-chemical properties or biological activities. Building quantitative relationships between molecular structures and properties, and explaining and predicting the properties and functions of the material from theory are very important to quantitative structure activity/property relationships (QSAR/QSPR). The key is to find the simple molecular structural descriptors which can efficiently reflect the molecular itself.(1) We built series simple descriptors according to molecular structural characterization of 20 substituted phenols. The molecular descriptors can not only represent substituents′differences, but also reflect the positions of substituents. The equations were established between molecular structure parameters and phenols acute toxicity for microtox, guppy, bluegill, trout, by using multivariate linear regression method. The results indicate that the models built by our methods show both better stability and stronger external predictive capability.(2) In this dissertation, support vector regression had been used for correlating and modeling the relationships between the geometrical parameters and environmental behaviors of polycyclic aromatic hydrocarbons. The prediction ability of the mathematical models obtained was somewhat better than that obtained by PLS regression.(3) We introduced an equilibrium electronegativity topology index mE(m=0,1) based on the structural characteristic bonding atom i and its mirror chemical surrounding of heteroatomic compounds, according to the adjacency matrix by defining equilibrium electronegativity of atoms and coloring atoms in molecular graph with equilibrium electronegativity,which was done through improvingδvi of atoms in the valence connection index,and defined a locating group parameterβ.And we also put the correlativity research on physicochemical properties of sulfides and aliphatic alcohols,alkoxy silicon chlorides.The results appeared that the physicochemical properties P of heteroatomic compounds can be described as follows:P=a0E+b1E+cβ+d.With this new method,calculation was convenient,and the results were accurate and had clear physical meanings.In addition,the novel topology index mE had unique to the molecular structures and excellent structural selectivity.The calculated results by the formula indicated that the average relative deviations between calculated values and experimental data were among the experimental deviations.