Description Of Chiral Center And Its Application To Enantioselectivity Prediction

The research of asymmetric reaction has an extremely high cost. Although thecombination of high-throughput methods and the combinatorial chemistry can accelerate theprocess of these investigations, a large number of experiments are still required. In order toobtain high enantiopure compounds in a time and cost-effective manner, in this thesis we builtseveral QSAR models to make enantioselectivity prediction as follows:1. Chiral Am index based on properties of ligands and its application.The absolute configuration of chiral center, called an S/R-like, was herein assignedrelative to the meaningful chemical properties of the ligands instead of atomic number, suchas size of the ligands. Then the chiral correction factor was obtained based on the absoluteconfiguraion of every chiral atom. At last, chiral Am index derived from2D topological indexwas obtained by introducing the correction factor. The new chiral indices combined withphysicochemical atomic stereodescriptors were used to successfully investigate theenantiopreference of Pseudomonas cepacia lipase (PCL) toward chiral primary alcohols.2. Virtual screening of a combinatorial library of enantioselective catalysts.Conformation-independent chirality codes, radial distribution function (RDF) codes, andindicator variables are implemented to represent catalysts in a combinatorial library for theasymmetric-hydrogen transfer to acetophenone. The catalysts which combine a metalic centrewith a chiral ligand have been evaluated in terms of both enantiomeric excess and yield. Acounterpropagation neural network (CPG NN) was trained with a small fraction of the libraryto predict the performance of catalysts, and applied to the virtual screening of the remaininglibrary. Selection of <20.8%of the virtual library with the highest predicted performanceenables to identify up to85.5%of the best catalysts. The approach illustrates achemoinformatic method to assist the optimization of resources for the screening ofenantioselective catalysts.3. Prediction of enantiomeric excesses in an asymmetric reaction.In order to assist enantioselective catalyst screening for asymmetric hydrogen transfer toacetophenone, the model of quantitative structure-activity relationship between theenantiomeric excesses (ee) of the products and the structures of the catalysts was constructed. In the presented study, several strategies were adopted to generate chirality codes ofcatalysts, and then the variables selection, from the initial pool of molecular descriptors, wasperformed by genetic algorithm. Finally, several stable mathematics models were obtained tomake prediction of ee.