Parallel Multiple Sequence Alignment Based On Hidden Markov Model

Compare is the common method of research in science, by conmare the study object to identify some feature on it.In the study of bioinformatics, comparing various biological sequence similarities is mandated by the sequence alignment. Sequence alignment is used to research of the domain protein identification, secondary structure prediction, gene identification, and molecular phylogenetic analysis. Multiple sequence alignment sometimes used to distinguish a set of sequences differences, but its main used to describe a set of sequences similarity relationship for concise understanding of a gene family features. With biological sequence datas surge increase, developed for a larger sequences alignment of parallel computation algorithm is an urgent task. This paper studies the multiple sequence alignment algorithms and its parallel algorithm in bioinformatics.The main contents and results are as follows listed:1. For the classic multiple sequence alignment algorithm, study the dynamic programming algorithm and CLUSTAL programming and HMM algorithm.Make comparison and assessment for the performance on several algorithms.2. Study of the current main parallel computing technology, select Cluster Workstations for this paper as a technical platform for parallel compute.Give a multiple sequences alignment algorithm based on the parallel Hidden Markov Model.3. Base on the algorithm of this paper bring forward, design and implement a parallel computing platform based on Windows operation system by Microsoft VisualC++.net.4. Using a group of several similar biological protein sequence data for testing algorithm, comparative analysis with the classical multiple sequence alignment method. The results show that based on parallel HMM multiple sequence alignment algorithms to solve protein multiple sequence alignment problems is effective, but there are also some problems. Finally, the paper discuss the foreground of parallel HMM algorithm sequence analysis.