Bundeswehr Institute of Microbiology, Munich, Germany
Bundeswehr Institute of Microbiology, Munich, Germany.
J Clin Microbiol. 2018 Apr 25;56(5). doi: 10.1128/JCM.01900-17. Print 2018 May.
Discrimination of highly pathogenic bacteria, such as , from closely related species based on molecular biological methods is challenging. We applied matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) to a collection of strains and close relatives in order to significantly improve the statistical confidence of identification results for this group of bacteria. Protein mass spectra of 189 verified and diverse strains of the group were generated using MALDI-TOF MS and subsequently analyzed with supervised and unsupervised statistical methods, such as shrinkage discriminant analysis (SDA) and principal-component analysis (PCA). We aimed at identifying specific biomarkers in the protein spectra of not present in closely related species. We could identify 7, 10, 18, and 14 -specific biomarker candidates that were absent in , , , and strains, respectively. Main spectra (MSP) of a defined collection of strains were compiled using the Bruker Biotyper software and added to an in-house reference library. Reevaluation of this library with 15 hitherto untested strains of and yielded improved score values. The strains were identified with score values between 2.33 and 2.55 using the in-house database, while the same strains were identified with scores between 1.94 and 2.37 using the commercial database, and no false-positive identifications occurred using the in-house database.
基于分子生物学方法区分高致病性细菌,如 ,与其密切相关的物种具有挑战性。我们应用基质辅助激光解吸电离飞行时间质谱(MALDI-TOF MS)对一组 菌株和近亲进行了研究,以显著提高该组细菌鉴定结果的统计置信度。使用 MALDI-TOF MS 生成了 189 个经验证且多样化的 组菌株的蛋白质质谱,并随后使用有监督和无监督统计方法(如收缩判别分析(SDA)和主成分分析(PCA))进行分析。我们旨在鉴定 蛋白谱中不存在于密切相关的 物种中的特定生物标志物。我们能够鉴定出 7、10、18 和 14 种特异性生物标志物候选物,它们分别不存在于 、 、 和 菌株中。使用 Bruker Biotyper 软件编译了一组定义的 菌株的主图谱(MSP),并将其添加到内部参考库中。使用 15 个以前未经测试的 和 菌株重新评估该库,提高了得分值。使用内部数据库对这些菌株进行鉴定,得分值在 2.33 到 2.55 之间,而使用商业数据库对相同的菌株进行鉴定,得分值在 1.94 到 2.37 之间,并且使用内部数据库不会出现假阳性鉴定。