Fatemi Mohammad Hossein, Gharaghani Sajjad
Department of Chemistry, Mazandaran University, Babolsar, Iran.
Bioorg Med Chem. 2007 Dec 15;15(24):7746-54. doi: 10.1016/j.bmc.2007.08.057. Epub 2007 Sep 1.
In this work some chemometrics methods were applied for modeling and prediction of the induction of apoptosis by 4-aryl-4-H-chromenes with descriptors calculated from the molecular structure alone. The genetic algorithm (GA) and stepwise multiple linear regression methods were used to select descriptors which are responsible for the apoptosis-inducing activity of these compounds. Then support vector machine (SVM), artificial neural network (ANN), and multiple linear regression (MLR) were utilized to construct the nonlinear and linear quantitative structure-activity relationship models. The obtained results using SVM were compared with ANN and MLR; it revealed that the GA-SVM model was much better than other models. The root-mean-square errors of the training set and the test set for GA-SVM model are 0.181, 0.241 and the correlation coefficients were 0.950, 0.924, respectively, and the obtained statistical parameters of cross validation test on GA-SVM model were Q(2)=0.71 and SRESS=0.345 which revealed the reliability of this model. The results were also compared with previous published model and indicate the superiority of the present GA-SVM model.
在本研究中，一些化学计量学方法被用于通过仅根据分子结构计算的描述符对4-芳基-4-H-色烯诱导细胞凋亡进行建模和预测。遗传算法（GA）和逐步多元线性回归方法被用于选择对这些化合物的凋亡诱导活性起作用的描述符。然后利用支持向量机（SVM）、人工神经网络（ANN）和多元线性回归（MLR）构建非线性和线性定量构效关系模型。将使用SVM获得的结果与ANN和MLR进行比较；结果表明GA-SVM模型比其他模型要好得多。GA-SVM模型训练集和测试集的均方根误差分别为0.181、0.241，相关系数分别为0.950、0.924，并且GA-SVM模型交叉验证测试获得的统计参数为Q(2)=0.71和SRESS=0.345，这表明该模型的可靠性。结果还与先前发表的模型进行了比较，表明了当前GA-SVM模型的优越性。
