论文部分内容阅读
使用比较分子场分析法建立了维甲类化合物抗致癌活性的三维构效关系模型。用交叉验证法和非交叉验证方法分别验证和建立的分子场模型有高的交叉验证回归系数（Ｒ２ｃｖ＝０．９０５），说明系列化合物分子周围空间场和静电场分布的差异与其生物活性差异间存在良好的相关性。使用这个模型预测在建立模型时没有包括进去的三个化合物的活性，预测值接近实验值，提示该模型具有较好的活性预测能力，可用来指导设计新的高活性维甲分子。化合物分子的最低能量构象未必是其活性构象。用得到的化合物最低能量构象进行ＣｏＭＦＡ研究，其交叉验证系数较低（Ｒ２ｃｖ＝０．４２０），不具有统计学意义。但将部分分子的单键微弱旋转，构象能变化控制在２千卡／摩尔之内，其它分子采取最低能量构象，则得到高的交叉验证系数。叠合的配体分子场模拟了配体分子周围的作用环境，反映受体结合部位与配体之间存在相互作用的基团和（或）原子的空间和静电性质，分子场模型在作为预测活性模板的同时也在一定程度上映射出受体结合部位的三维拓扑形状和理化特性。
A three-dimensional structure-activity-activity relationship model of anti-cancer activity of retinoids was established by comparative molecular field analysis. The cross-validated and established molecular field models with cross-validation and non-cross-validation methods, respectively, have high cross-validated regression coefficients (R2cv = 0.905) indicating that the difference between the spatial and electrostatic field distributions around the molecules of the compounds and their bioactivities There is a good correlation. This model was used to predict the activity of three compounds that were not included in the model establishment. The predicted value is close to the experimental value, which indicates that the model has good activity prediction ability and can be used to guide the design of new highly active retinoids. The lowest energy conformation of a compound molecule is not necessarily its active conformation. CoMFA studies with the lowest energy conformations of the compounds obtained showed a lower cross-validation coefficient (R2cv = 0.420), with no statistical significance. However, when the single bond of some molecules is weakly rotated, the conformational energy is controlled within 2 kcal / mol, and the other molecules adopt the lowest energy conformation to obtain a high cross-validation coefficient. The superimposed ligand molecular field simulates the surrounding environment of the ligand molecules, reflecting the space and electrostatic properties of the interacting groups and / or atoms of the receptor binding sites and ligands. The molecular field model is used as a prediction The active template also reflects the three-dimensional topological shape and physicochemical properties of the receptor binding site to a certain extent.