RNA二级结构预测的自适应鱼群算法模型  

Self-adaptive Artificial Fish Swarm Algorithm for RNA Secondary Structure Prediction

作　　者：何静媛[1] 邹东升[1] 何中市[1] 

机构地区：[1]重庆大学计算机学院,重庆400044

出　　处：《系统仿真学报》2010年第6期1370-1374,1396,共6页Journal of System Simulation

基　　金：重庆大学国家大学生创新性实验计划项目(CQUCX-G-2007-18)

摘　　要：基于人工鱼群算法的基本理论,提出了预测RNA二级结构预测的自适应人工鱼群算法模型。该模型对人工鱼群算法在处理组合优化问题时存在的缺陷进行了相应改进,在算法中加入的自适应思想既保证了算法的快速收敛,同时又有效地避免了盲目搜索。仿真实验中将本模型与基于粒子群算法的模型以及基于遗传模拟退火算法的模型进行了比较研究,结果表明,使用自适应鱼群算法进行RNA序列的二级结构预测能获得较理想的预测效果。较之另两种方法,鱼群算法能有效减少计算量、节省计算时间,特别当待测序列长度大于500时,鱼群算法在收敛速度上有着明显优势。A self-adaptive artificial fish swarm optimization algorithm for RNA secondary structure prediction was proposed, and this algorithm improves the traditional artificial fish swarm optimization algorithm by adding some self-adaptive idea. Simulation experiment shows that the self-adaptive artificial fish swarm algorithm can rapidly close to the global best result by avoiding the blindness of searching at the later stage. Compared with the particle swarm optimizer method and the genetic-simulated-annealing method, the algorithm not only can reach an acceptable prediction accuracy but also has much lower runtime cost, so it is a good choice to predict the secondary structure of a RNA molecule with more than 500 bases.

关 键 词：RNA二级结构 人工鱼群算法 自适应 组合优化 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] Q7[自动化与计算机技术—计算机科学与技术]