分子网络多靶标筛选的粒子群数值模拟法  

Multiple target optimal intervention using particle swarm optimization in molecular network

作　　者：周福家[1] 张宏伟[1] 李卫国[1] 

机构地区：[1]大连理工大学数学科学学院,大连116024

出　　处：《计算力学学报》2015年第2期269-273,279,共6页Chinese Journal of Computational Mechanics

基　　金：国家重点基础研究发展规划(2009CB918501)资助项目

摘　　要：当前对于疾病分子网络的治疗多采用多靶标药物治疗。多靶标药物治疗共分为两步,首先是筛选潜在的药物靶标,其次利用筛选出的药物靶标进行组合,进而得到有效的抑制方案。但现有算法耗时太多,本文使用花生四烯酸(Arachidonic acid)炎症代谢网络作为一个具体算例,使用改进并行的粒子群算法进行多靶标筛选优化,不仅能够得到药物靶标的有效组合,而且根据进程的个数可以使计算时间达到数倍的提高。The therapy of disease-related molecular network is usually based on the multiple target drug therapy. The multiple target drug therapy can be divided into two steps. First, filtering the potential drug targets. Second, making use of them to compose, in order to find the effective inhibition scheme. However, the present algorithm takes too much time. This article used the Arachidonic acid inflammation metabolic network as a concrete example, and also used the improved parallel particle swarm optimization to optimize the multiple targets. As a result,it can not only compose the drug targets effectively,but also save more time than the present algorithm according to the number of processes.

关 键 词：多靶标筛选 AA代谢网络 并行粒子群算法 疾病分子网络 

分 类 号：O241[理学—计算数学]