论文部分内容阅读
采用三维全息原子场作用矢量(three-dimensional holographic vector of atomic interaction field,3DHo VAIF)对31种磺胺类药物进行结构表征,分别采用多元线性回归(MLR)和偏最小二乘回归(PLS)建立磺胺类药物与其p Ka值的定量结构-性质相关(Quantitative Structure-Property Relationship,QSPR)模型.以R(建模相关系数),Rcv(交互检验相关系数),RMSEF(拟合误差)和RMSEP(预测误差)作为衡量模型预测能力的标准,其值分别为0.950,0.926,0.547,1.18(MLR);0.896,0.783,0.595,0.919(PLS).结果表明三维全息原子场作用矢量能较好地表征该类分子的结构信息,所建模型具有良好的稳定性和预测能力;同时也指出了sp3杂化的N原子与sp2杂化的S原子之间的静电作用是影响磺胺类药物p Ka值的重要因素.
31 sulfonamides were characterized by three-dimensional holographic vector of atomic interaction field (3DHo VAIF). Sulfonamides were established by multiple linear regression (MLR) and partial least squares regression (PLS) Quantitative Structure-Property Relationship (QSPR) model was used to determine the p Ka value of the drug and its p Ka value. The R (modeling correlation coefficient), Rcv (correlation coefficient of interaction test), RMSEF (fitting error) and RMSEP Error) as the standard to measure the predictive ability of the model, the values are 0.950,0.926,0.547,1.18 (MLR), 0.896,0.783,0.595,0.919 (PLS) respectively.The results show that the three-dimensional holographic atom field vector can characterize the It is also pointed out that the electrostatic interaction between the sp3 hybridized N atom and the sp2 hybridized S atom is an important factor affecting the pKa value of sulfonamides factor.