A Review of Biosequences Alignment, Matching, and Mining Based on GPU

Abstract: Sequence alignment, pattern matching, and mining are important cornerstones in bioinformatics, and they include identifying genome structure, protein function, and biological metabolic regulatory network. However, because it helps speed up the dealing process, the parallel sequential pattern recognition method has gained attention as data volume has increased. This review summarizes the GPU-based sequence alignment, pattern matching, and mining with the tools and their applications in bioinformatics. After giving an overview of the background, this review first introduces the concept and database of sequence alignment, pattern matching, and mining. Then, the basic architecture and parallel computing principle of GPU are briefly described. Next, the design of GPU-based algorithms and optimization strategies in sequence alignment, pattern matching, and mining are listed in detail. By comparing and analyzing the existing research, the summarization of the advantages and challenges of GPU application in bioinformatics are given. Finally, the future research direction is prospected, including the further development of the algorithm combined with machine learning and deep learning.