文章基本信息

标题：Structural Elucidation of Guaiane Sesquiterpenes from ¹³C Data using Generalized Regression Neural Network (GRNN) and Scatter Plots Methods
本地全文：下载
作者：Taye T. Alawode ; Kehinde O. Alawode
期刊名称：Annals. Computer Science Series
印刷版ISSN：1583-7165
电子版ISSN：2065-7471
出版年度：2017
卷号：15
期号：2
页码：79-88
出版社：Mirton Publishing House, Timisoara
摘要：This study seeks to elucidate of structures of unknown Guaiane sesquiterpenes from their 13C chemical shift values using Generalized Regression Neural Network (GRNN) and scatter plot methods. The 13C values for each of the fifteen (15) positions of the skeletons of the Guaiane sesquiterpenes were predicted using GRNN. From these predicted values, the substituents attached to each of the positions were predicted using GRNN and Scatter Plot methods. In predicting the skeletal 13C values, the 13C data of 116 Guaiane sesquiterpenes were used as input to GRNN while their corresponding data were used as the target data. The network was trained and simulated using twenty-five (25) test compounds. The best results were obtained at a spread constant of 5.0. In predicting the substituents on each position on the skeleton using the GRNN approach, the predicted 13C skeletal values of the test compounds were simulated following training of the GRNN using skeletal data of the 116 compounds as input data and their corresponding substituents (given codes) as the target data. The best results were obtained at a spread constant of 5.0. In the Scatter Plots method, graphs of codes of substituents for the 116 guaiane sesquiterpenes were plotted against the corresponding 13C chemical shift values of the skeletal Carbon to which they were attached. This gave the range of values over which each substituent may be obtained. The most likely substituent(s) for each position were selected. The degree of recognition of the test compounds (from both methods) ranged between 46.67 and 100%. Both methods gave similar recognition rates for the test compounds. GRNN and Scatter plots demonstrated great potential for use in the structural elucidation of unknown compounds from 13C values.
关键词：GRNN; Scatter plots; structural elucidation; 13C; Guaiane sesquiterpenes