基于多目标进化策略算法的DNA核酸编码设计  

A Multiobjective Evolution Strategy Algorithm for DNA Sequence Design

作　　者：张凯[1,2] 陈彬 许志伟 ZHANG Kai;Chen Bin;Xu Zhiwei(School of Computer Science and Technology,Wuhan University of Science and Technology,Wuhan 430065,China;Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System,Wuhan 430065,China)

机构地区：[1]武汉科技大学计算机科学与技术学院,武汉430065 [2]智能信息处理和实时工业系统湖北省重点实验室,武汉430065

出　　处：《电子与信息学报》2020年第6期1365-1373,共9页Journal of Electronics & Information Technology

基　　金：国家自然科学基金(61472293,61702383,61602328)。

摘　　要：设计高质量的核酸分子集合能有效提高DNA计算的可靠性、有效性和可求解问题的规模。DNA分子需要满足热力学约束、相似度约束、GC含量约束等多个相互冲突的目标函数,是典型的多目标优化问题。该文提出一种多目标进化策略(MOES)算法求解DNA分子序列设计问题,算法设计了随机碱基变异算子实现高效的局部搜索和全局搜索。改进的评价函数综合考虑了候选解的支配关系和冲突目标的平衡程度,选取符合DNA编码约束的核酸序列。实验结果证明,该文提出的算法具有高效的搜索效率和快速收敛能力,可以产生高质量的DNA序列集合,优于其他对比算法产生的DNA分子序列集合。It is important to design high-quality DNA sequences set,which can improve the reliability and efficiency of DNA computing.DNA sequence design problem is an multiobjective optimization problem that needs to satisfy multiple conflict objectives which are thermodynamic constraint,similarity constraint and GC content constraint simultaneously.A MultiObjective Evolutionary Strategy(MOES)is proposed to solve the DNA sequence design problem.The random base mutation operator is designed for exploration and exploitation the search space.The fitness function is improved for obtaining balanced similarity and H-measure objective functions.Some state-of-the-art approaches are chosen to evaluate the effectivity of proposed algorithm.The experiment results show that the proposed multiobjective evolution strategy algorithm obtains very promising DNA sequences and outperforms previous approaches.

关 键 词：DNA计算 DNA序列设计 多目标进化算法 进化策略 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]