Applications Of QSPR In Material Chemistry And Environmental Chemistry

Quantitative structure-property relationships (QSPR) studies are important research topics in chemoinformatics. They are used for the prediction of various physicochemical properties of compounds by statistical method and theoretical calculation. It is significant process to quickly select molecular descriptors and build a efficient model. In this paper the Heuristic Method (HM) and Best Multi-Linear Regression (BMLR) are used to select descriptors and build the linear model. To get a more efficient model than the linear model, the nonlinear model is built by Radial Basis Function Neural Networks (RBFNN) and support vector machine (SVM). The result is very satisfactory.Chapter 1 gives a description of the QSPR principle, research process and progress. In this section, we focus on describing modeling methods which are used in Chapter 2. At last we gave a review of the QSPR application.In Chapter 2, QSPR is applied to solving the recent chemical problem. A brief description of our studies was given as follows:A quantitative structure-property relationship model was developed to study the 25 polyvinylchloride (PVC) plasticizers'plasticization efficiency. HM was used both for pre-selecting molecular descriptors and for developing the linear model. The stability and validity of the model were evaluated by validation. The method which provided a good way to predict the plasticization efficiency of PVC plasticizers from their structures alone gave some insight into structural features related to PVC plasticizers.Quantitative structure-property relationship studies based on molecular descriptors for building models of a set of agricultural chemicals were presented. The octanol-water, soil-water partition coefficients (Kow, Koc) and bioconcentration factor (BCF) were correlated with five descriptors by HM and RBFNN. One biodegradation coefficient namely second-order rate coefficient (Kb) was used to classify chemicals by the biodegradation potential. This paper provided a method for testing and estimating the physicochemical property and biodegradation coefficient of new agricultural chemicals. It made people understand the relationship of agricultural chemicals and environment better.To understand the factors of cesium chelates'stability and predict stability constant value of a new macrocyclic compound, quantitative structure-property relationship studies based on molecular descriptors were presented for building models of cesium chelates. The stability constant, namely association constant (logK), was correlated with six descriptors by BMLR, RBFNN and Uniform Design Optimized Support Vector Machine (UDO-SVM). This method provided a novel method for testing and estimating association constants of the cesium chelates. It will be an experimental guide to find macrocyclic compounds which are used to detect and minimize the radiocesium pollution.