Research On Multiple Sequence Alignment Based On Adaptive Immune Genetic Algorithm

Sequence alignment is one of the base information disposal methods in bioinformatics research, which obtained abroad recognition and thorough study with advance of human genes plan, but at present, there is not an optimal algorithm of multiple sequence alignment.In recent years, excellent performances of genetic algorithm arouse attention of a lot of people and applied successfully in different domain which can solve optimize problems. How to amend research capacity and improve convergence speed of genetic algorithm, making it well applied in practical problems solving, it is a master problem that different country scholars research all along. Immune algorithm get attention and development freshly and possesses some characteristic distinguish from other algorithms. Combining immune algorithm with genetic algorithm, global research capability of genetic algorithm and local optimal capability of immune algorithm are all increased.This paper educes multiple sequence alignment based on adaptive immune genetic algorithm by studying multiple sequence alignment and analyzing trait of genetic algorithm,immune algorithm. In the algorithm, I adopting star-alignment algorithm to generate initial population, in this way, we may make the best of sequence's self information, which is better than insert gap into sequence randomly. It will spend more time in initial stages, but improving searching efficiency greatly in later alignment.Genetic algorithm is global convergent, but crossover operator and mutation operator are fixed relatively, so lacks alterable agility degree, which make genetic algorithm generating earliness phenomenon and getting into local extremum more easily. Therefore this paper adjusts crossover probability and mutation probability dynamically in course of evolution base on practice instance of population. Adaptive genetic algorithm keeps diversity of population,ensures astringency and advances optimization ability of algorithm. In addition, immune operator is applied to adaptive genetic algorithm, in precondition of reserve choiceness capability of original algorithm, restraining degeneration phenomenon in process of optimization much more availably.This paper presents material steps, proving global convergence by examples. I testify the combination is valid by theory analysis and experiment outcome.