基于基质辅助激光解析电离飞行时间质谱的B组链球菌血清型预测模型验证  

Validation of a prediction model for group B streptococcus serotypes based on matrix-assisted laser desorption ionization time of flight mass spectrometry

作　　者：黄莲芬[1] 陈光炼 高坎坎[1] 李姿娴 关小珊 钟华敏[1] 邓秋连 谢永强[1] 刘海英[1,2] Huang Lianfen;Chen Guanglian;Gao Kankan;Li Zixian;Guan Xiaoshan;Zhong Huamin;Deng Qiulian;Xie Yongqiang;Liu Haiying(Department of Laboratory Medicine,Guangzhou Women and Children's Medical Center,Guangzhou Medical University,Guangzhou 510120,China;Department of Laboratory Medicine,Guangzhou Medical University Brain Hospital,Guangzhou 510370,China)

机构地区：[1]广州医科大学广州市妇女儿童医疗中心检验科,510120 [2]广州医科大学附属脑科医院检验科,510370

出　　处：《中华生物医学工程杂志》2021年第2期134-141,共8页Chinese Journal of Biomedical Engineering

基　　金：广州市科技创新委员会项目(201804010447);广州市妇女儿童医疗中心院内课题(YIP-2019-051)。

摘　　要：目的验证基于基质辅助激光解析电离飞行时间质谱(MALDI-TOF MS)技术的B组链球菌(Group B streptococcus,GBS)在线血清型预测模型的分型效能,为临床GBS菌血清型快速分型提供参考。方法选取2015至2018年全国多中心新生儿侵袭感染分离GBS菌237株,均已经应用血清凝集试验结合PCR法确认血清型和多重PCR法多位点序列分型(MLST),包括Ⅰa型18株、Ⅰb型70株、Ⅲ型140株、Ⅴ型8株、Ⅵ型1株。运用乙醇灭活-甲酸(乙腈)提取法提取蛋白,用微生物质谱仪采集蛋白指纹图谱,将图谱以m/z和峰强度数据格式输入GBSTyper,以传统血清凝集法血清型为金标准,采用召回率、精确率、F值(精确率和召回率的调和平均值)来评价GBSTyper分型效果。结果GBSTyper对120株(50.63%)GBS菌血清型分型正确,各血清型召回率从高到低依次为Ⅲ型(71.43%)、Ⅴ型(37.50%)、Ⅰb型(21.43%)、Ⅰa型(11.11%),Ⅲ型精确率最高(86.96%),血清型分型精确率仅为3.57%~46.88%,Ⅲ型F值最高(78.43%),Ⅰa、Ⅰb、Ⅴ型等F值仅为6.52%~29.41%。结论GBSTyper血清型快速分型模型对Ⅲ型血清型菌株分型有效性较高,而对Ⅰa、Ⅰb、Ⅴ型分型有效性很低。该模型能快速识别高危Ⅲ型GBS,对于围产期新生儿GBS感染预防及患儿的临床管理有重要指导意义。Objective To verify the typing efficiency of an online prediction model for Group B streptococcus(GBS)serotypes based on matrix-assisted laser desorption ionization time of flight mass spectrometry(MALDI-TOF MS),so as to provide reference for rapid serotyping of GBS in clinical settings.Methods Included in this study were 237 strains of GBS strains isolated from infants with invasive infections in medical centers throughout China between 2015 and 2018.These GBS strains had been subjected to serotyping via agglutination test combined with PCR,and to multilocus sequence typing(MLST)via multiplex PCR,comprising 18 strains of typeⅠa,70 of typeⅠb,140 of typeⅢ,8 of typeⅤ,and 1 of typeⅥ.The GBS protein extraction was completed with ethanol inactivation-formic acid(acetonitrile)method,and the protein fingerprint was obtained with a microbial mass spectrometer.Then the mass spectrum data were input into the web-based tool GBS Typer in format of m/z and peak intensity.Using traditional latex agglutination serotyping as the gold standard,the recall,precision,F value(harmonic mean of precision and recall)were used to evaluate the typing efficiency of GBSTyper.Results GBSTyper classified 120 GBS strains(50.63%)correctly.The ranking of serotypes with recalls from high to low was typeⅢ(71.43%),typeⅤ(37.50%),typeⅠb(21.43%),and typeⅠa(11.11%).The serotype prediction yielded the highest accuracy for typeⅢ(86.96%)compared with only 3.57%to 46.88%for typesⅠa,Ⅰb,andⅤ,and highest F value for typeⅢ(78.43%)compared with only 6.52%to 29.41%for typesⅠa,Ⅰb,andⅤ.Conclusion The GBSTyper as a rapid serotyping model is effective for typeⅢbut fairly poor for typesⅠa,Ⅰb,andⅤGBS.As this model can rapidly identify the high-virulent typeⅢGBS,it provides important guidance for prevention and clinical management of perinatal GBS infection in the newborns.

关 键 词：无乳链球菌 光谱法 质量 基质辅助激光解吸电离 肽峰 预测模型 

分 类 号：R446.5[医药卫生—诊断学] R722.1[医药卫生—临床医学]