作　　者： ;

机构地区： 广东商学院数学与计算科学学院

出　　处： 《计算机工程与应用》 2007年第18期72-74,77,共4页

摘　　要： 生物序列motif识别问题是当今生物信息学面临的一个复杂问题,要设计一个能识别所有motif的方法几乎是不可能的。针对该问题,在免疫遗传算法中引入了统计估计,提高了motif识别的精度,根据个体的浓度和适应值概率。设计了免疫替换算子,有效地解决了种群的多样性问题,利用Gibbs Sampler算法生成种子,提高了免疫遗传算法的搜索速度,最后得到了一个基于免疫GA与Gibbs Sampler的生物序列motif识别算法,该算法充分发挥了免疫遗传算法和Gibbs Sampler算法的优越性,较好地解决了计算速度和计算精度之间的矛盾。实验表明,该算法是有效的。 Biological sequence motif detection is a complex problem in bioinformatics,it is barely possible to design a method with the ability of discovering all motifs in biological sequences.In the paper,in order to detecting motifs,the precision of motif detection has been increased by embedding the statistical estimate into immune genetic algorithm.The diversity of population has been solved by designing a substitution operator according to fitness and density probability of individual.The algorithm search speed has been improved by using Gibbs Sampler to breed seed.Finally,An approach based on immune GA and Gibbs Sampler for biological sequence motif detection has been constructed.The algorithm not only sufficiently exerts the advantages of the two algorithms,but also solves the contradiction between the computational speed and precision.The experiments show that the algorithm is effective.

关 键 词： 识别 生物信息学 免疫遗传算法 统计估计 抽样

领　　域： [自动化与计算机技术] [自动化与计算机技术]