新头孢菌素衍生物的定量构效关系研究

目的　研究一类新合成的抗流感嗜血杆菌头孢菌素衍生物的定量构效关系。方法　应用半经验量子化学方法和误差反向传播人工神经网络的方法。结果　系统计算了4 8个头孢菌素类化合物的1 8个分子描述符,从中筛选出7个描述符。随机挑选4 3个化合物作为训练样本集,5个化合物作为测试样本集,检验结果表明模型具有较高的精度(检验样本最大相对误差小于5 %)。结论　头孢菌素化合物的LUMO能量、疏水性、C8原子及C3 位取代基的净电荷是影响其抗菌活性的主要因素,所建立的定量构效关系模型能够有效地进行体外抗流感嗜血杆菌活性与头孢菌素类化合物分子描述符的相关性分析。     

QSAR studies of antibacterial ricinoleic acid derivatives

A series of ricinoleic acid derivatives has been synthesized and tested for antibacterial activity with respect to four standard strains. Dibromoricinoleic acid (DBRA) showed high activity comparable with that of the reference drug ciprofloxacin. QSARs between various physicochemical indices and the antibacterial activity of a training set including 12 compounds were analyzed. The topological parameter, the valence second-order molecular connectivity index (²χ^v), and the electronic parameter of total energy (TE) proved to be important for the antibacterial activity of compounds studied. The proposed QSAR models were validated using the leave-one-out procedure. The validity of these models was confirmed by predicting the activity of a set of three compounds (not present in the training set). __________ Translated from Khimiko-Farmatsevticheskii Zhurnal, Vol. 41, No. 3, pp. 16–21, March, 2007.