基于量子化学方法发现新型酪氨酸激酶抑制剂

Based on the quantum chemical method to find new tyrosine kinase inhibitors

采用量子化学中的密度泛函理论方法,在B3LYP/6-31++G水平下,系统计算了11种亚苄丙二腈类衍生物的量子化学参数,并通过回归分析方法,构建了二维定量构效模型,分析了影响活性抑制的主要因素,并建立了亚苄丙二腈衍生物与抑制酪氨酸激酶活性之间的定量构效关系方程,研究结果表明:该类化合物分子的分子总能量ET与疏水性参数log P对抑制活性的影响最大,且疏水性能越强,分子的活性抑制能力越高。在此基础上,使用留一法交叉验证了模型的预测能力,结果表明,模型的回归系数和留一法交叉验证系数分别为0.796和0.7291,表明模型具有较好的预测能力,可以用于预测此类化合物的活性。基于结构相似性,设计了4种新型的亚苄丙二腈类衍生物分子,在相同水平下计算其量子化学参数,并预测其活性,结果表明这些新型新型酪氨酸激酶抑制剂均具有较好的活性,研究结果为进一步设计性能更好的酪氨酸激酶抑制剂提供了理论参考。

The quantum chemical parameters of 11 kinds of benzylidenemalononitrile derivatives were calculated at the B3LYP / 6-31 ＋＋G level by density fimctional theory, the two-dimensional quantitative structure-activity relationship （QSAR） was constructed by regression analysis, analyzed the main factors of influencing the inhibition of activity and constructed quantitative structure-activity relationship between benzimoniummalononitrile derivative and inhibiting tyrosine kinase activity. The results showed that, the molecular total energy ET and the hydrophobicity parameter logP had the greatest influence on the inhibition activity, and the stronger of hydrophobic property, the higher of the molecule＇s activity inhibition. On this basis, the prediction ability of the model was verified by using the leave-out method. The result showed that the regression coefficients and cross-validation coefficients of the model are 0.796 and 0.7291, respectively, which indicated that the model has good prediction ability, could be used to predict the activity of such compounds. Based on structural similarity, 4 kinds of benzylidenemalononitrile derivatives were designed and calculated the quantum chemical parameters at the same level and its activity was predicted. These new results showed that the new tyrosine kinase inhibitors have better activity, which provide a theoretical reference for the further design of better tyrosine kinase inhibitors.