Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 May 9;10:947.
doi: 10.3389/fmicb.2019.00947. eCollection 2019.

Comparative Analysis of Tools and Approaches for Source Tracking Listeria monocytogenes in a Food Facility Using Whole-Genome Sequence Data

Free PMC article

Comparative Analysis of Tools and Approaches for Source Tracking Listeria monocytogenes in a Food Facility Using Whole-Genome Sequence Data

Balamurugan Jagadeesan et al. Front Microbiol. .
Free PMC article


As WGS is increasingly used by food industry to characterize pathogen isolates, users are challenged by the variety of analysis approaches available, ranging from methods that require extensive bioinformatics expertise to commercial software packages. This study aimed to assess the impact of analysis pipelines (i.e., different hqSNP pipelines, a cg/wgMLST pipeline) and the reference genome selection on analysis results (i.e., hqSNP and allelic differences as well as tree topologies) and conclusion drawn. For these comparisons, whole genome sequences were obtained for 40 Listeria monocytogenes isolates collected over 18 years from a cold-smoked salmon facility and 2 other isolates obtained from different facilities as part of academic research activities; WGS data were analyzed with three hqSNP pipelines and two MLST pipelines. After initial clustering using a k-mer based approach, hqSNP pipelines were run using two types of reference genomes: (i) closely related closed genomes ("closed references") and (ii) high-quality de novo assemblies of the dataset isolates ("draft references"). All hqSNP pipelines identified similar hqSNP difference ranges among isolates in a given cluster; use of different reference genomes showed minimal impacts on hqSNP differences identified between isolate pairs. Allelic differences obtained by wgMLST showed similar ranges as hqSNP differences among isolates in a given cluster; cgMLST consistently showed fewer differences than wgMLST. However, phylogenetic trees and dendrograms, obtained based on hqSNP and cg/wgMLST data, did show some incongruences, typically linked to clades supported by low bootstrap values in the trees. When a hqSNP cutoff was used to classify isolates as "related" or "unrelated," use of different pipelines yielded a considerable number of discordances; this finding supports that cut-off values are valuable to provide a starting point for an investigation, but supporting and epidemiological evidence should be used to interpret WGS data. Overall, our data suggest that cgMLST-based data analyses provide for appropriate subtype differentiation and can be used without the need for preliminary data analyses (e.g., k-mer based clustering) or external closed reference genomes, simplifying data analyses needs. hqSNP or wgMLST analyses can be performed on the isolate clusters identified by cgMLST to increase the precision on determining the genomic similarity between isolates.

Keywords: CFSAN pipeline; Listeria monocytogenes (L. monocytogenes); Lyve-SET; core genome MLST (cgMLST); high quality single nucleotide polymorphism (hqSNP); smoked salmon; whole genome MLST (wgMLST); whole genome sequence (WGS).


Maximum parsimony tree based on k-mer-based SNP analysis. The tree was built using kSNP3 with the core SNPs identified among the set of 42 isolates in the study dataset plus 140 L. monocytogenes closed genomes downloaded from the NCBI RefSeq database. Lineages (I, II, and III), the three clusters (1, 2, and 3) and sub-clusters (3a and 3b), as well as the unclustered isolate are annotated. Percentages of consensus clustering agreement across up to 100 equally parsimonious trees are shown for the clusters identified in this study and main nodes representing the L. monocytogenes lineages. The tree was midpoint rooted.
Maximum likelihood phylogenetic tree based on hqSNP analysis using the BioNumerics (BN) pipeline. The tree was constructed with RAxML using core hqSNPs identified within cluster 1 (A), sub-cluster 3a (B), and sub-cluster 3b (C). Bootstrap values greater than 70% are shown above the branches. Clades within (sub-) clusters are shown with the hqSNP ranges identified with the three hqSNP methods (CFSAN, Lyve-SET, and BN).

Similar articles

See all similar articles

Cited by 2 articles


    1. Allard M. W., Bell R., Ferreira C. M., Gonzalez-Escalona N., Hoffmann M., Muruvanda T., et al. (2018). Genomics of foodborne pathogens for microbial food safety. Curr. Opin. Biotechnol. 49 224–229. 10.1016/j.copbio.2017.11.002 - DOI - PubMed
    1. Allard M. W., Strain E., Melka D., Bunning K., Musser S. M., Brown E. W., et al. (2016). Practical value of food pathogen traceability through building a whole-genome sequencing network and database. J. Clin. Microbiol. 54 1975–1983. 10.1128/JCM.00081-16 - DOI - PMC - PubMed
    1. Bankevich A., Nurk S., Antipov D., Gurevich A. A., Dvorkin M., Kulikov A. S., et al. (2012). SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J. Comput. Biol. 19 455–477. 10.1089/cmb.2012.0021 - DOI - PMC - PubMed
    1. Bolger A. M., Lohse M., Usadel B. (2014). Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30 2114–2120. 10.1093/bioinformatics/btu170 - DOI - PMC - PubMed
    1. Bridier A., Sanchez-Vizuete P., Guilbaud M., Piard J. C., Naitali M., Briandet R. (2015). Biofilm-associated persistence of food-borne pathogens. Food Microbiol. 45 167–178. 10.1016/ - DOI - PubMed