In our QSAR study, pharmacophore identification and biological activity estimation of 80 methanone derivatives were performed with the Electron Conformation Genetic Algorithm approach. Using the geometric, thermodynamic and topological properties of molecules from the data obtained from quantum chemical calculations in the HF/3-21 G basis set, the Electron Conformational Matrices of Congruity were generated by the EMRE software. Taking into account the pharmacophores atoms, 804 parameters were prepared. The nonlinear least squares optimization technique and genetic algorithm were used to determine the variables affecting the biological activity values for the calculation of the biological activity values of the studied molecules. 4D-QSAR approach the EC-GA method, that ensures pharmacophore detection, variable selection and quantitative bioactivity prediction, is used to calculate biological activity values of methanone derivatives. The model for the training and test sets attained by the optimum 8 parameters gave highly satisfactory results with R²_training = 0.834, q² = 0.768 and SE_training = 0.075, q²_ext1 = 0.875, q²_ext2 = 0.839, q²_ext3 = 0.764, ccc_tr = 0.908, ccc_test = 0.929 and ccc_all = 0.920. The interaction between the studied molecules and the live cancer protein which ID is 2H80 at DockingServer was examined to find the activity of the molecules examined in molecular placement calculations.

中文翻译：

---

电子构象遗传算法的甲烷酮衍生物分子对接和4D-QSAR模型

在我们的QSAR研究中，使用电子构象遗传算法方法对80种甲酮衍生物进行了药效团鉴定和生物活性估计。利用从HF / 3-21 G基集中的量子化学计算获得的数据，使用分子的几何，热力学和拓扑性质，通过EMRE软件生成了电子构象矩阵。考虑到药效团原子，制备了804个参数。使用非线性最小二乘最优化技术和遗传算法确定影响生物活性值的变量，以计算所研究分子的生物活性值。4D-QSAR采用EC-GA方法，可确保药效团检测，变量选择和定量生物活性预测，用于计算甲酮衍生物的生物活性值。通过优化的8个参数获得的训练和测试集模型给出了令人满意的结果R ²_训练 = 0.834，q ²  = 0.768和SE_训练 = 0.075，q ² _ext1  = 0.875，q ² _ext2  = 0.839，q ² _ext3  = 0.764，ccc _tr  = 0.908，ccc _test  = 0.929和ccc _all  = 0.920。在DockingServer上检查了研究的分子与ID为2H80的活癌蛋白之间的相互作用，以发现分子放置计算中所检查分子的活性。