Epidemiological typing of methicillin-resistant Staphylococcus aureus strains by Fourier transform infrared spectroscopy
Epidemiological typing of methicillin-resistant Staphylococcus aureus strains by Fourier transform infrared spectroscopy
复制标题
DOI：
10.1016/j.mimet.2006.12.022
复制
发表时间：
2007-04-01
影响因子：
2.2
通讯作者：
Ismail, Ashraf A.
中科院分区：
文献类型：
--
作者：
Amiali, Nassim M.;Mulvey, Michael R.;Ismail, Ashraf A.
需要一种快速简单的分型系统来控制流行性耐甲氧西林金黄色葡萄球菌（MRSA）的传播，MRSA是目前加拿大医院中最广泛的多重耐药医院病原体之一。傅里叶变换红外（FTIR）光谱分型85株代表5株流行性加拿大MRSA（CMRSA）。将在Que-Bact（R）通用培养基2号上生长的全细胞的光谱指纹转化为一阶导数峰高归一化文件，并通过视觉和奇异值分解（SVD）进行检查。通过主成分分析、自组织映射和K近邻监督聚类分析对光谱区域进行了区分。在视觉识别的区域中，1070-1050和1155-1137 cm（-1）分别被发现适合于区分CMRSA-4和CMRSA-2，而CMRSA-1，CMRSA-3和CMRSA-5在1123-1094 cm（-1）区域中各自表现出独特的光谱轮廓。1123-1094、1174-1154和2904-2864 cm（-1）的组合将5个CMRSA分开，PCA的正确分类率为84.6%。还使用SVD选择的区域1096-1066、1118-1090和2914-2880 cm（-1）获得了五个聚类，其中基于PCA分数图的视觉检查的正确分类率为87.8%，基于监督聚类分析的正确分类率为97%。这些结果表明，FTIR光谱作为一种快速（1小时）和简单的方法，在临床环境中MRSA菌株分型和监测具有相当大的潜力。(c)2007 Elsevier B. V.保留所有权利。
A rapid and simple typing system is needed for controlling the spread of epidemic methicillin-resistant Staphylococcus aureus (MRSA), currently one of the most widespread multi-resistant nosocomial pathogens in Canadian hospitals. Fourier transform infrared (FTIR) spectroscopy was used to subtype 85 isolates representing five strains of epidemic Canadian MRSA (CMRSA). Spectral fingerprints of whole cells grown on Que-Bact (R) Universal Medium No. 2 were transformed to first derivative peak-height normalized files and examined visually and by singular-value decomposition (SVD). Distinguishing spectral regions were processed by principal component analysis (PCA), self-organizing map and K-nearest neighbor supervised cluster analysis. Among the visually identified regions, 1070-1050 and 1155-1137 cm(-1) were found suitable for discrimination of CMRSA-4 and CMRSA-2 respectively, while CMRSA-1, CMRSA-3, and CMRSA-5 each exhibited distinctive spectral profiles in the 1123-1094 cm(-1) region. The combination, 1123-1094, 1174-1154 and 2904-2864 cm(-1) separated the five CMRSA with 84.6% correct classification by PCA. Five clusters were also obtained using the SVD-selected regions 1096-1066, 1118-1090 and 2914-2880 cm(-1), with 87.8% correct classification based on visual examination of the PCA scores plot and 97% based on supervised cluster analysis. These results demonstrate that FTIR spectroscopy has considerable potential as a rapid (1-hour) and simple method for MRSA strain typing and monitoring in clinical settings. (c) 2007 Elsevier B.V. All rights reserved.