Research And Design Of Hardware Acceleration For Sequence Alignment Algorithm

The main task of sequence alignment is to determine the degree of similarity between two or more sequences.This work is an important part of gene sequencing and a very central issue in bioinformatics.Sequence alignment requires a large amount of computing resources,and current alignment algorithms are difficult to meet the massive biological information data processing requirements.In this paper,the research on accelerated processing of the sequence alignment algorithm BWA-MEM is carried out.Firstly,this paper studies the BWA-MEM algorithm,analyzes its algorithm flow and main computing strategies,and analyzes the performance of its different computing stages to determine that the Seed Extension stage is computationally intensive.Then,the feasibility of hardware acceleration is analyzed for the smith-waterman algorithm.Secondly,a heterogeneous computing hardware platform is set up.This paper uses the wavefront characteristics of Smith-Waterman algorithm to propose a hardware acceleration method for Smith-Waterman algorithm.The design of the hardware accelerator is based on the Open CL language and a high-level comprehensive design flow.The hardware-accelerated design of the Smith-Waterman algorithm on the Intel FPGA hardware platform is analyzed and its acceleration effect is analyzed and discussed.Finally,in order to reduce the redundant time consumed by data transmission and calling the kernel in the hardware acceleration platform,this paper uses the data integration optimization design strategy to further improve the completed Smith-Waterman algorithm hardware acceleration design and software optimization for BWA-MEM.And then,this paper achieves accelerated optimization of BWA-MEM heterogeneous computing.The experimental results show that the hardware acceleration platform and hardware acceleration method proposed in this paper can achieve the acceleration of Smith-Waterman algorithm.In the range of sequence length less than 150,the acceleration ratio of its operation time can be up to 4 times.Based on smith-waterman hardware accelerator and bwa-mem software optimization,the maximum speed-up ratio of 1.34 can be achieved under ideal conditions for the average operation time of Extend in the Seed Extension phase of bwa-mem.The research results show that the hardware design and acceleration method of sequence alignment proposed in this paper can accelerate the sequence alignment algorithm,and it is expected to be applied to the actual sequence alignment application.