摘要
为了指导设计合成更高生物活性的噁唑烷酮类化合物,本文利用Hyperchem 7和Chemoffice 2004软件包计算了60个噁唑烷酮类化合物的物理化学、电子结构、拓扑指数等参数,利用偏最小二乘、穷举回归和混沌遗传算法训练的人工神经网络方法建立了定量构效关系模型,根据模型详细讨论了这些参数对抗菌活性的影响,结果表明：当该类抗菌化合物具有较大的水合能和分子表面积时有利于提高抗菌活性；当具有较低的最低空轨道能、结合能、分子拓扑指数和较小的分子体积时对提高其抗菌活性有利。
Oxazolidinone antibiotics and its analogues were calculated using the Hyperchem 7 and ChemOffice 2004 to design and synthesize higher biological activity of oxazolidiuoe and analogues. Quantitative structure-activity relationship studies have been approached based on Physieochemieal, electronic structure, topology indices etc. parameters, using partial least squares, exhaustion regression and artificial neural networks trained by chaos map aided genetic algorithm. The antibacterial activities of oxazolidinone and analogues will increase due to the larger hydration energy and surface area; the lower energy of the lowest unoccupied molecular orbital, binding energy, volume and molecular topological index can enhance the antibacterial activities.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2006年第7期663-667,共5页
Computers and Applied Chemistry
基金
河南省杰出青年科学基金(项目编号:0612002600)
化学生物传感与计量学国家重点实验室(湖南大学)开放课题
河南省高校青年骨干教师计划项目
