论文部分内容阅读
应用启发式算法(HM)和基因表达式编程方法(GEP)建立了31种磺胺类药物pKa值的定量构效关系模型。用ChemOffice2004软件进行化合物的结构输入,利用半经验方法进行分子结构优化,在CODDESA软件中计算出组成、拓扑、几何、电子和量子化学参数,并用启发式方法筛选出4个相关参数,在此基础上运用多元线性回归和基因表达式编程方法建立QSPR模型。两种方法均得到了较好的结果,HM和GEP的的相关系数分别为0.90和0.95。两种QSPR模型在新药研究中可以预测化合物的pKa值,将为新药研究提供理论指导。
A quantitative QSAR model was established for the pKa values of 31 sulfonamides using heuristic algorithm (HM) and gene expression programming (GEP). The structures of compounds were input by using ChemOffice2004 software. The molecular structure was optimized by semi-empirical method. The composition, topological, geometric, electronic and quantum chemical parameters were calculated by CODDESA software. Four relevant parameters were screened by heuristic method. On the use of multiple linear regression and gene expression programming method to establish QSPR model. Both methods gave good results, and the correlation coefficients of HM and GEP were 0.90 and 0.95, respectively. The two QSPR models predict the pKa of a compound in new drug research and will provide theoretical guidance for new drug research.