期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2017
卷号：17
期号：10
页码：231-237
出版社：International Journal of Computer Science and Network Security
摘要：Smith-Waterman (S-W) algorithm is the perfect sequence alignment method for the biological database but practically this algorithm lacks pace due to high computational complexity. FASTA, BLAST and other heuristics approaches are faster in computations but less accurate. Volume and length variation of sequences require restructuring the database. Acceleration of Smith-Waterman algorithm on proper modern hardware brings perfection and accuracy. This paper presents a high-performance sequence alignment algorithm implemented on Kepler’s architecture graphic processor unit. This new implementation is improved version having reduced memory accesses to eliminate bandwidth congestion. The implementation is performed on Kepler’s architecture graphics processing unit on which the performance was raised to 51 Giga Cells updates per second GCPUS which is 138.3% increase than the previous implementation on GTX275 GPU. In this implementation protein database is converted into equal length sequence sets on advanced GPU. By this workload is distributed among GPU microprocessor threads. This results in improved implementation than previous implementations.
关键词：Smith Waterman; SwissProt; Proteins; Sequencing; Alignment; GCPUS; FASTA
