TABLE 5.

Characteristics of the available analytical methods for the detection and enumeration of potentially enteropathogenic Vibrio spp. in seafood and their contribution to risk assessment.

	Culture‐based detection methods	Culture‐based quantification methods	Conventional typing methods	WGS (from isolates)	Metagenomics (targeted and untargeted)
Strengths	A validated ISO standard is available for Vp, Vv and Vc Provide comparable data on prevalence for exposure assessment Provide isolates for further characterisation, including detection of pathogenicity markers Technically easy and affordable	An ISO standard is available for Vp Provide data on contamination levels for exposure assessment Support the acquisition of data for modelling (including DR models, growth/inactivation models, effect of climate change, efficacy of control measures, etc.) May provide isolates for further characterisation Can provide both data on proportion of samples with pathogenic strains and proportion of pathogenic strains on total for that species within a sample	Allow strain discrimination Allow easy identification of epidemic clones Support basic outbreak investigation Support basic investigation of strain introduction in an environment Affordable compared to WGS	Robust for isolates characterisation High‐level phylogenetic resolution for outbreak investigation and trace‐back Provide information for source attribution and identification of transmission routes/pathways Data can be analysed (and re‐analysed) for different aims Data sharing is technically easy	Provide microbial population data (composition and variation, resistome, etc.) Bypass cultivation or isolation, therefore applicable also to non‐cultivable, fastidious, or slow‐growing Vibrio spp. of interest, even achieving the whole genome of some organisms
Weakness	Performance comparison between standard methods is not available In some methods pathogenicity characterisation (e.g. tdh/trh of Vp, serogroup of Vc) is optional Random isolation of a limited number of colonies may lead to underestimation of the proportion of samples containing pathogenic strains Time consuming	No interlaboratory validation for currently available methods No ISO standard for quantitative analysis of Vv or Vc Not systematically coupled to strain isolation for further characterisation Time consuming and labour intensive	Lack resolution to discriminate closely related strains Rapid obsolescence of techniques and databases due to advent of WGS Time consuming and labour intensive	Metadata sharing is complex (legal framework, ethical issues, etc.) Insufficient standardisation of data analysis Data analysis requires specialised personnel Not yet affordable for all potential users	Lack of sensitivity for detection species or sub‐groups within species of interest Absence of isolates for further analysis Absence of phenotypic testing Insufficient standardisation of data analysis Not yet applicable for untargeted detection of species of PH interest and monitoring purposes Data analysis requires trained personnel Not yet affordable for all potential users
Opportunities	Easy to combine with molecular screening methods (e.g. conventional, real‐time and viability PCR) and with detection of new molecular targets Easy to couple with genomic characterisation of isolates Support acquisition of prevalence data also in challenging areas (e.g. low‐income countries, areas with basic laboratories)	May support the definition of quantitative microbiological criteria or trade standards	Support acquisition typing data also in challenging areas (e.g. low‐income countries, areas with basic laboratories)	Potential for routine use in both diagnostic and food control User‐friendly bioinformatic pipelines can extend use and acquisition of data from non‐specialised laboratories Support hazard identification through subtype discrimination (beneficial for a more targeted risk assessment) Identification of new sequences for microorganism detection and characterisation (prediction of virulence, host association, stress resistance, etc.) in association with other omics Potential for new approaches integrating these data into risk assessment (hazard identification, characterisation and exposure assessment) to support decision making	May detect new or re‐emerging infectious agents Potential to study microorganisms dynamics within complex microbial communities
Threats	The lack of VBNC detection may lead to underestimation of prevalence and to underestimation of exposure	The lack of VBNC detection may lead to underestimation of contamination levels and to underestimation of exposure	The lack of discriminative power may lead to false epidemiological associations	The insufficient standardisation in data analysis may lead to inconsistent results	May report signals from non‐viable organisms, leading to an overall inaccurate interpretation of results

Abbreviations: DR, dose response; ISO, International Organization for Standardization; PCR, polymerase chain reaction; PH, public health; TDH, thermostable direct haemolysin; TRH, TDH‐related haemolysin; VBNC, viable but non‐culturable; Vc, V. cholerae; Vp, V. parahaemolyticus; Vv, V. vulnificus; WGS, whole genome sequencing.