| 乌药化学成分结构参数与色谱保留时间的关系 |
| |
| 引用本文： | 堵锡华,王超.乌药化学成分结构参数与色谱保留时间的关系[J].浙江大学学报(理学版),2018,45(6):721-727. |
| |
| 作者姓名： | 堵锡华 王超 |
| |
| 作者单位： | 徐州工程学院 化学化工学院, 江苏 徐州 221018 |
| |
| 基金项目： | 国家自然科学基金资助项目（21472071，21473081）. |
| |
| 摘 要： | 乌药含有呋喃倍半萜、黄酮、生物碱、挥发油等多种化学成分，具有抗氧化、抗肿瘤及抑菌等活性.为研究乌药化学成分的性质，计算了乌药化学成分的分子连接性指数和电性拓扑状态指数，并筛选分子连接性指数中的0X，5X和5Xc，以及电性拓扑状态指数中的E1，E2和E3，将这6种结构指数作为神经网络的输入层变量，色谱保留时间作为输出层变量，做回归分析，采用6∶5∶1的神经网络结构，构建了预测能力较强的预测模型.该模型总相关系数rt为0.994 0，训练集相关系数为0.992 9，测试集相关系数为0.997 0，验证集相关系数为0.997 9，利用该模型计算得到的保留时间预测值与实验值吻合度较好，相对平均误差为2.66%.结果表明，乌药化学成分的色谱保留时间与6种结构参数之间具有良好的非线性关系.
|
| 关 键 词： | 乌药 化学成分 色谱保留时间 分子连接性指数 电性拓扑状态指数 神经网络 |
| 收稿时间： | 2017-11-02 |
Correlation between structural parameters of Lindera Aggregata's chemical compositions and the retention time of chromatogram |
| |
| DU Xihua,WANG Chao.Correlation between structural parameters of Lindera Aggregata's chemical compositions and the retention time of chromatogram[J].Journal of Zhejiang University(Sciences Edition),2018,45(6):721-727. |
| |
| Authors: | DU Xihua WANG Chao |
| |
| Affiliation: | School of Chemistry and Chemical Engineering, Xuzhou Institute of Technology, Xuzhou 221018, Jiangsu Province, China |
| |
| Abstract: | Lindera aggregata consists of many chemical compositions, such as furan sesquiterpene, flavonoid, alkaloid and volatile oil, and has the features of anti-oxidant, anti-tumor and anti-bacterial. In order to further study the features of Lindera aggregate's chemical compositions, the molecular connectivity indices and electrical topological state indices of Lindera aggregata's chemical compositions were calculated. The molecular connectivity indices 0X, 5X and 5Xc, as well as the electrical topological state indices E1, E2 and E3 were selected. Then, the regression method was employed to analyze the six indices and the retention time of chromatogram. We used this six indices as input variables of a neural network and the retention time of chromatogram as the output variable. Meanwhile, the 6:5:1 network structure was adopted to establish a satisfying neural network model with good predictive ability. The total correlation coefficient rt of the model was 0.994 0,the correlation coefficient r1 of training set was 0.992 9, the correlation coefficient r2 of test set was 0.997 0 and the correlation coefficient r3 of validation set was 0.997 9. The predicted values and the experiment values were relatively close, and the mean relative error was 2.66%. The results show that there exists a good nonlinear relationship between the retention time of chromatogram and the six structural parameters. |
| |
| Keywords: | Lindera aggregata chemical compositions retention time of chromatogram molecular connectivity index electrical topological state index neural network |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《浙江大学学报(理学版)》浏览原始摘要信息 |
|
点击此处可从《浙江大学学报(理学版)》下载全文 |
