病原宏基因组高通量测序生物信息学分析质量评价的研究现状与思考  

The Research Status and Considerations in Quality Evaluation of Bioinformatics Analysis of Metagenomic Next-Generation Sequencing

作　　者：刘东来[1,2,3] 关文达 吴林寰[5] 王浩[1,2] 杨子峰[4] 许四宏 LIU Dong-lai;GUAN Wen-da;WU Lin-huan;WANG Hao;YANG Zi-feng;XU Si-hong(National Institutes for Food and Drug Control;NMPA Key Laboratory for Quality Research and Evaluation of Medical Devices;NMPA Key Laboratory for Quality Research and Evaluation of In Vitro Diagnostics;Guangzhou Institute of Respiratory Health,the First Affiliated Hospital of Guangzhou Medical University;National Microbiology Data Center,Institute of Microbiology,Chinese Academy of Sciences)

机构地区：[1]中国食品药品检定研究院 [2]国家药品监督管理局医疗器械质量研究与评价重点实验室 [3]国家药品监督管理局体外诊断试剂质量研究与评价重点实验室 [4]广州医科大学附属第一医院广州呼吸健康研究院 [5]中国科学院微生物研究所国家微生物科学数据中心

出　　处：《中国食品药品监管》2024年第5期34-41,共8页China Food & Drug Administration Magazine

基　　金：国家重点研发计划(2022YFC2305300);中国食品药品检定研究院关键技术研究基金(GJJS-2022-2-2)。

摘　　要：病原宏基因组高通量测序(mNGS)技术已成为感染病原学诊断的新工具,由实验操作(湿实验)和生物信息学分析(干实验)两部分组成。干实验由算法和数据库构成,其功能是对湿实验产生的测序数据进行分析处理后输出检测结果。干实验的性能受到测序数据中复杂多变的干扰因素的影响,包括临床样本中大量的人源核酸、试剂与耗材携带的微生物核酸、采样与湿实验引入的环境微生物核酸、数据库中基因组质量不均一或不同物种基因组之间的相似性过高导致的错误比对与注释,以及算法与参数差异对分类鉴定的影响等。上述干扰因素可能来自mNGS检测各个环节,不仅可能导致干实验输出错误的物种鉴定和微生物检测结果,也给干实验的质量控制与评价带来较大挑战。本文综述了mNGS干实验质量控制的关键问题以及关于质量评价方法的思考。Metagenomic next-generation sequencing(mNGS)has become a new tool for the diagnosis of infectious pathogens,consisting of two parts:experimental procedures(wet bench)and bioinformatics analysis(dry bench).The dry bench,comprising algorithms and databases,aims to output detection results after analyzing and processing the sequencing data generated by the wet bench.The performance of the dry bench is affected by complex and variable interference factors in sequencing data,including a large number of human nucleic acids in clinical samples,microbial nucleic acids carried by reagents and consumables,environmental microbial nucleic acids introduced by sampling and experimentation,incorrect alignment and annotation caused by heterogeneous genome quality in databases or excessive similarity between genomes of different species,and the influence of algorithm and parameter differences on classification and identification.These interference factors may come from various steps of the mNGS workflow,potentially leading to incorrect species identification and microbial detection results from dry bench,and posing great challenges to its quality control and evaluation.This paper reviews the key issues of quality control of mNGS dry bench and discusses thoughts on quality evaluation methods.

关 键 词：病原宏基因组高通量测序 生物信息学分析 质量评价 数字参考品 

分 类 号：R95[医药卫生—药学]