Blast-Parallel: The parallelizing implementation of sequence alignment algorithms based on Hadoop platform

The sequence alignment is a basic method for processing the information in Bioinformatics, it has a great significance for finding the function and the structure of nucleic acids and protein sequences and the information of evolution. This paper briefly describes the relevant issues of sequence alignment and the most common local sequence alignment algorithms, Blast algorithm. At present, the Blast algorithm which provided by NCBI or stand-alone can not meet the actual demand for the flood of biological data, this paper achieves the Blast-Parallel algorithm by further improvement based on the Hadoop-Blast algorithm. Through serial experiments of the stand-alone Blast algorithm and parallelizing experiments of the Hadoop-Blast algorithm and the Blast-Parallel algorithm based on Hadoop platform, results show that the Blast algorithm has significantly higher execution efficiency after the parallelization, and the matching speed of the Blast-Parallel algorithm which has been improved can achieve 1~1.5 times of the Hadoop-Blast algorithm.