Exploiting Multi-level Parallelism for Homology Search using General Purpose Processors

New biological experimental techniques are continuing to generate large amounts of data using DNA, RNA, human genome and protein sequences. The quantity and quality of data from these experiments makes analyses of their results very time-consuming, expensive and impractical. Searching on DNA and protein databases using sequence comparison algorithms has become one of the most powerful techniques to better understand the functionality of particular DNA, RNA, genome, or protein sequence. This paper presents a technique to effectively combine fine and coarse grain parallelism using general purpose processors for sequence homology database searches. The results show that the classic Smith-Waterman sequence alignment algorithm achieves super linear performance with proper scheduling and multi-level parallel computing at no additional cost