A Design Of Short Gene Sequence Alignment Acceleration System Based On High Performance Hash Table

In November 2016,the State Council issued the "Thirteenth Five-Year Plan" for the development of national strategic emerging industries,and in December 2016,the national development and Reform Commission issued the "Thirteenth Five-Year Plan" for the development of biological industry,all of which proposed to popularize the application of gene sequencing and promote the development of gene sequencing industry to more benefit people's livelihood.In clinical applications,gene sequencing can be used for screening for birth defects and early screening of tumors.In animal and plant systems,green pesticides,veterinary drugs,and animal vaccines based on genome information of pests have been widely used in people's production and life.The concept of "high-throughput sequencing",mentioned many times in the "Thirteenth Five-Year Plan" industry plan,represents a gene sequencing technology with high data throughput and requiring huge computing power.High-throughput sequencing technology covers the second,third,and even fourth-generation gene sequencing technologies.It is the mainstream direction of current gene sequencing technology.The development of highthroughput and low-cost gene sequencing instruments and other medical equipment has national strategic significance.Gene sequence alignment is a part of high-throughput sequencing technology and is also the part that consumes the most computing power and sequencing time.The scattered,disordered,and numerous test gene sequences are compared with the reference genome,and then a large amount of genetic information is analyzed to reconstruct the test genome.The research direction of this subject is to reduce the time and cost of gene sequence alignment,and then to reduce the time and cost of gene sequencing.This paper presents a hardware acceleration system for gene sequence alignment with a special structure,which discards the arm + FPGA heterogeneous platform used in traditional hardware development and uses the X86 + FPGA heterogeneous platform based on the Open CL specification.This structure can greatly improve the data processing capability of software platform and data transmission speed between two platforms.The X86 platform works well with the FPGA platform to perform highly parallel computing.In this paper,a set of complete gene sequence alignment scheme is established.The speed and accuracy of sequence index are guaranteed by using the gene sequence preprocessing scheme based on hash index,and the scheme with global alignment and local alignment is innovatively used to improve the speed and accuracy of gene sequence alignment.The system implements a gene sequence preprocessing system(HASH)and a data transmission network of heterogeneous platform,and integrates a gene sequence global alignment system(MBV)and a gene sequence fine alignment system(SW).This design supports Whole Genome Sequencing and its working clock reaches 160 Mhz.The alignment speed of a single Kernel can reach 62 million sequence alignments per second,and the scheme with 5 Kernels can reach 180 million times per second.