Table 3.
Examplesa of published high throughput sequencing based approaches to detection of foodborne pathogens for industrial and control purposes.
| Target organisms | Type of matrix and material sequenced | Aim | Sequencing approach and technology | Data output | Bioinformatics approach to data analysis | References |
|---|---|---|---|---|---|---|
| BACTERIA | ||||||
| Salmonella and microbial community | Naturally contaminated tomatoes (4 samples, 4 conditions) | Pathogen detection (Salmonella) and characterization of the microbial community after pre-enrichment and enrichment | Shotgun metagenomics, MiSeq (Illumina) 151PEb | Raw data not described. After assembly: 1.5 × 107 sequences of variable length (mean = 210 bp), comprising a total of 2.6 Gbp | Taxonomic classification by reference-based analysis in MG-RAST. Pathogen detection by reference-based mapping to a database of Salmonella genomes | Ottesen et al., 2013 |
| Salmonella and microbial community | Naturally contaminated cilantro (coriander leaves). 91 samples included in metataxonomic study, 12 samples in shotgun metagenome study | Pathogen detection and quantification (Salmonella) and characterization of the microbial community before and after enrichment | Metataxonomics by 16S rDNA amplicon sequencing on GS FLX (Roche) and shotgun metagenome sequencing on MiSeq (75PE) with 1 to 6 samples per run | GS FLX raw data not described. Total number of shotgun reads ranged from 6.8 × 105 to 5.5 × 106 per sample depending on number of samples sequenced in one run | Taxonomic classification by reference-based analysis using the RDP classifier. Pathogen detection by reference-based mapping to genome sequences of cultured isolates of Salmonella from the same study | Jarvis et al., 2015 |
| Listeria monocytogenes and microbial community | Naturally contaminated ice cream. 3 enrichments with four replicates and 13 time points included in metataxonomic study, 6 samples included in shotgun metagenome study | Pathogen detection (Listeria monocytogenes) and characterization of the microbial community before, during and after enrichment | Metataxonomics based on 16S rDNA amplicon sequencing on MiSeq (300PE), and shotgun metagenome sequencing on MiSeq (PE151) | Metataxonomic data not described. Total number of shotgun reads ranged from 1.5 × 107 to 8.2 × 107 per sample | Taxonomic classification by reference-based analysis using Resphera Insight and subtyping of L. monocytogenes 16S rDNA fragments by mapping in PYNAST. Pathogen detection and shotgun metagenome analyses with Genius bioinformatics software for identification to species, subspecies and strain level by reference-based mapping to L. monocytogenes genomes | Ottesen et al., 2016 |
| STECc and microbial community | Bagged spinach: spiked and unspiked samples, altogether 16 samples | Pathogen detection and quantification of STECc, virulence gene detection and characterization of the microbial community before, during and after enrichment | Shotgun metagenome sequencing on MiSeq (250PE) | Total number of shotgun reads ranged from 8.0 × 106 to 1.5 × 107 per sample | Pathogen detection by reference-based mapping to virulence and serotype genes and STEC genomes. Microbial community analysis by k-mer analyses against bacterial genomes for identification of bacterial species, phylogroup or serovar | Leonard et al., 2015 |
| STEC and microbial community | Bagged spinach: 36 spiked samples (low level) | Pathogen detection and quantification (STEC), virulence gene detection and characterize the microbial community before, during and after enrichment | Shotgun metagenome sequencing on MiSeq (250PE) | Total number of shotgun reads ranged from 1.2 × 106 to 5.2 × 106 per sample | Pathogen detection by reference-based analyses targeting STEC, molecular serotypes, virulence genes and SNPs complemented with phylogenetic analyses. Microbial community analysis using a k-mer approach for identification of bacterial species, phylogroup or serovar | Leonard et al., 2016 |
| VIRUSES | ||||||
| Viral community | Lettuce: 42 field grown samples and 54 retail samples | Pathogen detection and quantification, and characterization of viral community | RNAseq and shotgun metagenome sequencing, complemented with 16S rDNA metataxonomics to verify absence of bacterial DNA on HiSeq2500 (Illumina; 100PE) | Total number of reads approximately 1.5 × 109 (approximately 2.3 to 3.4 × 107 reads per sample) | Assembly of reads followed by mapping to viral reference genomes for taxonomic classification, including detection of putative human pathogens | Aw et al., 2016 |
| Viral community | Fresh parsley plants irrigated with fecally tainted water + water samples and negative controls | Pathogen detection and quantification, and characterization of viral community | RNAseq on MiSeq (300PE) | From 1.2 × 106 (irrigated parsley), to 8.1 × 106 (negative control plants) reads per sample | Assembly of reads followed by mapping to viral reference genomes for taxonomic classification, complemented with phylogenetic analysis | Fernandez-Cassi et al., 2017 |
| Norovirus | Commercial shellfish | Pathogen detection and typing to assess diversity | RNAseq on HiSeq (100PE) and metataxonomics (capsid VP1 gene) on MiSeq (300PE) | Approximately 2 × 107 reads per sample (RNAseq) and from 2 × 105 to 4 × 105 reads per sample (metataxonomics) | Mapping of reads to viral reference genomes for taxonomic classification and genotyping | Imamura et al., 2016b |
| Norovirus | Farmed oysters | Pathogen detection and typing to assess the performance of a commercially applied decontamination system | Metataxonomics (capsid VP1 gene) on MiSeq (300PE) | From 2 × 105 to 4 × 105 reads per sample | Assembly and mapping to norovirus reference genomes for genotyping | Imamura et al., 2016a |
a
The examples are further described and discussed in the main text of this review.
b
151PE refers to the use of paired-end sequencing with a read length of 151 bp.
c
STEC, shiga-toxin producing E. coli.