【摘要】 在对4-[3-（吡啶-4-基）-3-取代苯基丙烯酰]吗啉类杀菌剂进行合理分类的基础上,应用遗传函数算法（GFA）分别建立优选描述符与化合物对2种病菌靶标杀菌活性（pEC50）的二维定量结构-活性关系（2D-QSAR）模型,并优选了模型所需的最佳描述符数量。所建模型都通过了交叉验证、随意检验、外部预测检验等显著性检验,r2、r2-CV、r2-pred均大于0.950,因而具有良好的预测可靠性。计算结果表明:化合物对立枯丝核菌（Rhizoctonia solani Kuhn.）的杀菌活性主要受S_ssCH2和Vm等描述符的影响;而对掘氏疫霉菌（Phytophthora drechsler Tucker）的杀菌活性主要受Atype_C4,Sr,Area和S_dO等描述符的影响。更多还原

【Abstract】 2-Dimensional quantitative structure-activity relationship （2D-QSAR） analysis of novel 4-[3-（pyrid- 4-yl）-3-substituted phenyl acryloyl] morpholine were investigated with selected descriptors based on the genetic function approximation （GFA） for their fungicidal activities （pEC50） against two types of pathogen drone, respectively. The compounds were rationally classified into training and test set using methods described by Oprea et al. The number of descriptors adequate in the model was optimized. The significance of the models was validated by cross-validation, randomization test and external test set prediction, indicating good predictive abilities of the two best models with values of r2, r2-CV and r2-pred exceeding 0.950. The results showed that the activity against Rhizoctonia solani Kuhn. was governed by the descriptors of S_ssCH2 and Vm; the activity against Phytophthora drechsler Tucker was controlled by the descriptors of Atype_C4, Sr, Area and S_dO. The findings in this study can provide academic guidance of structural design and synthesis of novel effective fungi- cides including pyridine and acryloyl morpholine derivatives.更多还原