原子电距矢量用于雄甾酮类衍生物核磁共振碳谱的定量构谱关系研究

Atomic Electronegative Distance Vector as Applied to QSSR Study for Estimation and Prediction of ^(13)CNMR Chemical Shift in Androsterones

提出以5元素构建原子电性距离矢量,描述雄甾酮化合物中不同等价碳原子的化学环境。对9个雄甾酮分子171个等价碳原子的13CNMR化学位移建立了多元线性回归模型。模型的复相关系数为R=0 9917,标准偏差为SD=5 5384。采用留一法交互检验的结果为RCV2=0 9284,SDCV=11 4866。结果表明模型具有良好的估计能力和稳定性。

A novel atomic electronegative distance vector (AEDV) is developed to express the chemical environment of various chemically equivalent carbon atoms in androsterones. A five-parameter model with linear relationship of chemical shift for 171 equivalent carbon atoms of 9 androsterones is created using multiple linear regression.The correlation coefficients (R) and standard deviation (SD) of molecular modeling are 0.9917 and 5.5384,respectively.To validate the prediction capabilities of these QSSR models for the external prediction set,the leave-one-out (LOO) cross-validation (CV) technique is applied,giving the correlation coefficient R_(CV)~2=0.9284 and standard deviation SD_(CV)=11.4866.The results show that AEDV is an excellent topological index with satisfactory estimation stability and favorable predictive ability.