摘要
为提高蛋白质折叠结构的预测精度,提出了一种融合改进量子遗传算法及局部搜索策略的蛋白质折叠结构预测方法.该方法在传统的量子遗传算法算法基础上引入动态调整量子门旋转角步长机制以及量子变异操作,从而提高算法的优化性能.局部搜索策略按照一定规则对量子遗传算法的优化结果进行局部结构变换,这种结构变换只需通过移动较少的节点就可以实现,能够有效提高算法的优化效率.计算机仿真实验表明,该算法能够获得较优的蛋白质折叠结构预测结果.
In order to increase the accuracy of predictions of protein folding,a predictive strategy using an improved quantum genetic algorithm with a local search was developed.In this method,step lengths for dynamically adjusting the angle of the quantum gate and quantum mutation operators were introduced,therefore high performance optimization was achieved.Local structural transformation was carried out in the local search;this was based on rules for optimizing results from the quantum genetic algorithm.Because this structural transformation can be derived by moving only a few vertices,the efficiency of optimization increased.Simulation results showed the method effectively improves the predictive accuracy of protein folding in comparison with other methods.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2010年第1期92-97,共6页
Journal of Harbin Engineering University
基金
国家自然科学基金资助项目(60873037)
黑龙江省教育厅科技基金资助项目(11511118)
关键词
生物信息学
蛋白质折叠
量子遗传算法
局部搜索
优化
bioinformatics
protein folding
quantum genetic algorithm
local search
optimization
