Research Of Rapid Microbe Detection Computational Methods Based On NGS

Biomedicine is moving into big data era.The development of next generation sequencing is changing researching methods of biological fields.Microbe detection is aimed at detecting the pathogen microbes in metagenomics sample.“Unknown” bio-threat is problematic because its characteristic of sudden occurrence,limitation of site and equipment etc.Microbe detection methods based on next generation sequencing is becoming into an important method of preventing “unknown” bio-threat for advantages of high accuracy and short duration.These methods is of great help for disease control and prevention,drug invention etc.It is a hot topic of biomedicine research.This paper analyzes recently published microbe detection calculation model based on next generation sequencing.To solve the problem that not all existed methods can be evaluated synthetically,we proposes Computational Analysis of Microbiome Identification Benchmark(CAMI benchmark),then use github to realize automatic evaluation process for detection methods.After that we put forward Adjustable Microbe Detection Algorithm(AMDA)based on pre-proposed sensitivity input requirement,evaluate it using CAMI benchmark.The result showes that in original mode,ADMA can realize detection speed as fast as Readscan and control the sensitivity error down to 3%.Eventually,we come up with microbe detection platform based on mobile sequencing(MDPMS)using FPGA and high performance computing platform.The experiments show that for 120 Gb sequencing data from HiSeq2500,MDPMS subtractes the running duration from 4 months to 1 week using FPGA and 2 days using high-performance computing platform.