遗传算法模拟生物多序列比对之初始化的改进  

Improvement of initialization for multiple sequence alignment by genetic algorithm

作　　者：李满枝[1] 王凯华[1] 王洪涛[1] 修兴强[1] 

机构地区：[1]海南师范大学数学与统计学院,海南海口571158

出　　处：《辽宁工程技术大学学报（自然科学版）》2017年第6期662-667,共6页Journal of Liaoning Technical University (Natural Science)

基　　金：海南省自然科学基金(20151003);国家自然科学基金(71361008);海南省重点项目(ZDXM20130080)

摘　　要：针对应用遗传算法求解生物多序列比对问题的初始化进行了改进.初始种群是遗传算法构造的一个关键部分,本文根据多序列比对的生物特性,在初始种群中的个体中插入连续空位,优化初始种群的个体质量;并在初始种群中加入一定比例的在线比对工具MAFFT优质种子,优化初始种群的整体质量.通过数值模拟实例结果显示,经过这两个优化处理可以生成更高质量的初始种群,得到更好的比对结果,提高多序列比对的计算效率.This paper used genetic algorithm to simulate biological multiple sequence alignment problem, and improved the initial population greatly. As the initial population is the most critical part of the genetic algorithm, this paper added continuous gaps in the individuals of initial population with the biological properties which can optimize its individual quality, and then added an appropriate proportion MAFFT excellent seed in initial population which can optimize its overall quality. The optimization can obtain better quality initial population efficiency of multiple sequence alignment is improved. experimental results show that combination of those two and get higher accuracy alignment, and the computational

关 键 词：多序列比对 遗传算法 初始化 连续空位 MAFFT种子 

分 类 号：O242.1[理学—计算数学]