喹啉类生物碱的^(13)C-NMR波谱模拟

Modeling Spectrum Relationship of ^(13)C-NMR Simulationfor of Quinoline Alkaloids

应用一种反映分子局部微环境描述子——原子电性相互作用矢量(vector of atomic electronegative interaction,AEIV)和原子杂化状态指数(atomic hybridation state index,AHSI)对喹啉类化合物的13种分子中的129个^(13)C-NMR谱建模模拟,应用多元线性回归方法得到定量结构波谱关系模型的复相关系数(R_(MM1))为0.988,标准偏差(SD_(MM1))为5.317.采用留一法交互检验结果R_(CV1)为0.987,SD_(CV1)为5.630.随机抽出两部分分子进行检验,得到的相关系数R_(MM2)为0.993,R_(MM3)为0.987.结果表明,使用AEIV和AHSI所建模型具有相当的预测能力和稳定性.

Atomic electronegativity interaction vector （AEIV） and atomic hybridation state index （AHSI） were employed for quantitative structure spectroscopy relationship modeling of ^13C NMR chemical shifs of 129 carbon atoms in thirteen molecules of quinoline alkaloids. Through multiple linear regression （MLR）, a model with the correlation coefficient （RMM,） of 0. 988 and the standard deviation （SDMM]） of 5. 317 could be achieved. Leave-one-out （LOO） method and cross-validation （CV） were combined to test the prediction ability of the model and the satisfied results with correlation coefficient （RCV1） of 0. 987 and standard deviations （SDCV1） of 5. 630 could be obtained, respectively. Furthermore, to test the validity of this model, two parts of the molecules were selected randomly as the test sets, with the correlation coefficients （RMM2and RMM3） of 0. 993 and 0. 987, respectively. The results suggest that the model based on AEIV and AHSI can be used to predict ^13C NMR chemical shifts of quinoline with satisfactory stability and predictability.