National Reference Laboratory for Listeria monocytogenes, German Federal Institute for Risk Assessment, Department of Biological Safety, Berlin, Germany.
Institute of Biology, Freie Universität Berlin, Department of Biology, Chemistry and Pharmacy, Berlin, Germany.
Microb Genom. 2021 Jan;7(1). doi: 10.1099/mgen.0.000491. Epub 2020 Dec 4.
Where classical epidemiology has proven to be inadequate for surveillance and control of foodborne pathogens, molecular epidemiology, using genomic typing methods, can add value. However, the analysis of whole genome sequencing (WGS) data varies widely and is not yet fully harmonised. We used genomic data on 494 isolates from ready-to-eat food products and food processing environments deposited in the strain collection of the German National Reference Laboratory to compare various procedures for WGS data analysis and to evaluate compatibility of results. Two different core genome multilocus sequence typing (cgMLST) schemes, different reference genomes in single nucleotide polymorphism (SNP) analysis and commercial as well as open-source software were compared. Correlation of allele distances from the different cgMLST approaches was high, ranging from 0.97 to 1, and unified thresholds yielded higher clustering concordance than scheme-specific thresholds. The number of detected SNP differences could be increased up to a factor of 3.9 using a specific reference genome compared with a general one. Additionally, specific reference genomes improved comparability of SNP analysis results obtained using different software tools. The use of a closed or a draft specific reference genome did not make a difference. The harmonisation of WGS data analysis will finally guarantee seamless data exchange, but, in the meantime, knowledge on threshold values that lead to comparable clustering of isolates by different methods may improve communication between laboratories. We therefore established a translation code between commonly applied cgMLST and SNP methods based on optimised clustering concordances. This code can work as a first filter to identify WGS-based typing matches resulting from different methods, which opens up a new perspective for data exchange and thereby accelerates time-critical analyses, such as in outbreak investigations.
在针对食源性致病菌的监测和控制方面，传统流行病学已被证明存在局限性，而利用基因组分型方法的分子流行病学则可以提供更多的价值。然而，全基因组测序（WGS）数据分析的方法差异较大，尚未完全实现标准化。我们使用了德国国家参考实验室菌株库中储存的 494 株来自即食食品和食品加工环境的分离株的基因组数据，对各种 WGS 数据分析程序进行了比较，并评估了结果的兼容性。我们比较了两种不同的核心基因组多位点序列分型（cgMLST）方案、单核苷酸多态性（SNP）分析中的不同参考基因组以及商业和开源软件。不同 cgMLST 方法的等位基因距离相关性很高，范围在 0.97 到 1 之间，统一的阈值比特定方案的阈值产生更高的聚类一致性。与使用通用参考基因组相比，使用特定参考基因组可将检测到的 SNP 差异数量增加多达 3.9 倍。此外，特定的参考基因组还可以提高使用不同软件工具获得的 SNP 分析结果的可比性。使用封闭或草案特定参考基因组并没有产生差异。WGS 数据分析的标准化最终将确保无缝的数据交换，但与此同时，了解导致不同方法聚类分离的相似阈值值可能会加强实验室之间的沟通。因此，我们根据优化的聚类一致性，在常用的 cgMLST 和 SNP 方法之间建立了一个翻译代码。这个代码可以作为一个初步的筛选器，用于识别来自不同方法的基于 WGS 的分型匹配，从而为数据交换开辟了新的视角，并加速了时间关键型分析，例如暴发调查。
