改进二进制量子粒子群算法在蛋白质折叠中的应用  

Improved binary quantum-behaved particle swarm optimization to protein folding

作　　者：赵晶[1,2] 孙俊[1] 须文波[1] 

机构地区：[1]江南大学物联网工程学院,江苏无锡214122 [2]山东轻工业学院信息学院,济南250353

出　　处：《计算机应用研究》2011年第9期3381-3383,共3页Application Research of Computers

基　　金：国家自然科学基金资助项目(60703106;60474030)

摘　　要：为了改善已有二维HP模型蛋白质折叠算法容易陷入局部最优、找不到理论最低能量构象的缺点,提出一种基于变异算子的改进二进制量子粒子群算法。采用二进制编码蛋白质序列,提出变异策略,并采用惩罚因子避免出现蛋白质重叠,最后将该算法应用于蛋白质序列进行测试。测试结果表明,改进算法能够找到更优的结果,算法具有一定的实用性和有效性。The existed protein folding algorithms in hydrophobic-polar model（HP model） are easily being trapped in local optima and can not obtain the minimum energy of protein folding conformation.To overcome the disadvantages,this paper proposed an improved binary quantum-behaved particle swarm optimization algorithm based on mutation operator.In the novel algorithm,introduced the binary coding to code amino acid sequence.Then proposed the mutation strategy to improve the premature phenomena.It adopted the punishing factor to avoid the overlapped protein folding.Tested some benchmark sequences to the proposed algorithm.The results of experiment show that the proposed technique can find the more excellent minimum energy of protein folding conformation than other algorithms.The proposed algorithm is practical and effective.

关 键 词：量子粒子群算法 二进制 变异 蛋白质折叠 二维HP模型 蛋白质序列 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]