Use of Whole Genome Sequencing (WGS) of food-borne pathogens for public health protection

...............................IV. OVERALL CONCLUSIONSWGS is a powerful tool that can be applied to a wide range of public health and foodsafety applications. Both SNP-calling and gene-by-gene comparison may be validapproaches for the different public health aims of WGS, as long as stable and comparable analytical procedures are used. Draft genomes can be assembled, using relatively modest computing power and widely available software, from data generated by commercially available high-throughput next generation sequencing (NGS) platforms. These include relatively inexpensive bench top instruments. Such draft genomes are suitable for the majority of practical applications in food safety. There are a number of ways in which this technology can be implemented in a food safety setting, including (i) installing the instruments in a food safety environment and (ii) outsourcing to commercial or other public health laboratories. The choice between these models will depend on local issues of throughput and available resources. Further work is required to fully implement WGS in routine application. This includes: (i) the adoption of appropriate quality assurance/quality control (QA/QC) measures; (ii) the development and harmonisation of SOPs; (iii) the establishment of database infrastructure; and (iv) the generation and dissemination of appropriate sets of genomic data. Reference datasets should comprise not only representative complete (i.e. finished) genomes, that can be used for reference mapping if required, but also a larger number of draft genomes with accompanying catalogues of allelic diversity data, required for gene-by-gene approaches. These databases must be open access and widely available, although sensitive epidemiological data will have to remain available only to the competent authorities. Further work is also needed to link these data (i) with previous isolate characterization schemes and nomenclatures, although these may have to be revised in the light of the new insights available, but also (ii) to the phenotypic properties of the isolates from which the genomic data are obtained. As the standardization of data generation, analysis, and storage is key, transnational organizations should aim to facilitate the implementation and integration of WGS across health sectors, working towards the goal of a ‘one health’ approach. It is clear that currently only a few organizations are investing in WGS approaches for realtime analysis of food-borne and other pathogens but many institutes are exploring the utility of these approaches through research collaborations or ‘in house’ using bench-top machines. There are various factors that will hinder more widespread, routine use of these technologies, including funding for machines, IT infrastructure and impact on current workforce. From a bioinformatics viewpoint, access to necessary bioinformatics knowledge, Overall conclusions42 EFSA Scientific Colloquium 20, Parma, 16-17 June 2014 bioinformatics staff, skills and infrastructure for data storage and management (both locally and centrally) are key issues. The need to share knowledge of laboratory and bioinformatics approaches was recognised as necessary to expedite use of WGS. Some sequencing protocols are publicly available but knowledge sharing between bioinformatics groups is informal at best and frequent forums/meetings would be greatly helpful. The greater challenges that lie ahead for the bioinformatics analysis of WGS applied to global public health are perhaps less of a technical IT nature, as technology and computer science is likely to continue to provide the solutions to handle the ever-growing data wave of microbial genomes. The generation and implementation of standards for data storage and sharing are clearly critical needs, since this will drive the data analysis development and will facilitate the harmonisation of WGS data analytical approaches. In terms of data analysis, a few models have emerged but we are still in the early days and, although the current models of data analysis should start to be used and their results compared, novel approaches could still emerge as the sequencing technology keeps evolving. Models for international data sharing resources were discussed including a single, central resource for sequence storage, management and analysis compared with a system of federated databases underpinned by common ontologies and APIs. Concerns were expressed over the scalability of the current global databases (e.g NCBI and EBI) in their current formats to meet the potential tsunami of data likely to emerge in coming years and also whether the considerable funding required for a single central resource would be found. A network of federated, perhaps national, databases may have advantages in terms of resilience but adoption of common APIs and ontologies would be essential. There are currently many developments in different countries to establish resources for centralised analysis of microbial genomes for public health and research purposes. It may be that, as the best approaches emerge, they will be most widely adopted but we should endeavour to ensure that there is interoperability between systems as they are designed and not as an afterthought. The discussions highlighted national differences in attitudes to data release and also differences between attitudes in public health and commercial groups. EFSA Scientific Colloquium 20, Parma, 16-17 June 2014 43 In the meantime, however, there are urgent needs that can help move global approaches forward: X Formal structures for multi-disciplinary collaboration, knowledge sharing and training in laboratory sequencing procedures and bioinformatics approaches is likely to expedite emergence and wider use of best practice. GMI has led some important initiatives but is currently unfunded. X Wider development and adoption of ontologies and APIs will facilitate data sharing and integration across multiple resources. This is essential to ensure best value gained from data generated. X It is recommended that the major structures leading the sharing of raw data (NCBI SRA and EBI ENA) continue to be adequately funded and supported, as they remain the major driver for this change by providing access to the baseline data for stakeholders in this field, from bioinformaticians to clinicians. The most effective use of this technology is not yet obvious and this was discussed extensively during the colloquium. A major difference between WGS and the classical typing methods is that WGS allows all genes to be included in the analysis, instead of a well-defined subset of genes or variable intergenic regions. Therefore, the analysis of WGS data will yield new types of insight. As it is conceptually new, the legal and official systems are not yet adapted to the large-scale application of WGS to support food safety policies. At present it is not clear whether new legislation is required for optimal utilization of WGS, or if the existing legal framework is sufficient. Not only the legal power of evidence supplied by WGS needs to be defined, but also issues such as the determination of quality standards. The need for cooperation is at the same time the Achilles’ heel of WGS. There are several impediments for the free sharing of data. While the deposition of the microbial genomic sequence data in databases for public access beyond the control of the owner of the data is common practice, releasing sensitive epidemiological data will not easily become a routine procedure. Legal obstacles are to be expected and a careful balance must be struck between the desirable complete openness from a food safety point of view and the privacy and related concerns that necessitate confidentiality. Possibly a standard for encryption may need to be developed, to allow exchange of data to be limited to authorized parties only. Ignoring these issues is likely to considerably delay the successful large-scale implementation of WGS for public health at international level. The overall conclusion of the EFSA Colloquium on whole genome sequencing is that the application of WGS for food safety and public health in general requires a paradigm Overall conclusions44 EFSA Scientific Colloquium 20, Parma, 16-17 June 2014 change and the development of new methodologies for processing these data. Since information obtained by WGS can only be utilized to fully benefit food safety and public health if all parties involved use procedures aimed at generating compatible results, international and inter-sector cooperation are crucial. To exploit the enormous potential of WGS, implementation in the EU must be initiated without delay. Recommendations Based on the considerations above the participants of the Colloquium recommended that in general: X Educational interdisciplinary programmes should be established to foster a more in-depth understanding of WGS allowing epidemiologists, bioinformaticians, microbiologists and food safety experts to process, integrate and correctly interpret WGS and epidemiological data for the benefit of food safety and public health. X Given that rapid technical developments can be expected, there is no need for stringent standardization of WGS methodologies, but quality metrics need to be defined. X Databases should be established to contain WGS data for public health, standards should be agreed upon for the curation of these data and minimal demands on the associated epidemiological data should be decided upon. X The legal consequences of the large-scale application of WGS should be reviewed and proposals should be made for any changes in food safety rules and regulations that may be needed. X All parties involved should strive to harmonize terminology and data reporting in plain language accounts aimed at policy makers and other non-specialist readers. X Software and data analysis tools should be developed that are at the same time transparent and easy to use for all staff involved in WGS data processing. X Reference datasets should comprise both finished complete genomes and draft genomes with catalogues of allelic diversity data.EFSA Scientific Colloquium 20, Parma, 16-17 June 2014 45 Specific recommendations for the EU: X EFSA and ECDC should assume a leading role within the EU framework to stimulate, steer and coordinate efforts for the application of WGS across health sectors to further food safety and protection of public health. X EU stakeholders in public health should work together with EU Member States’ competent bodies in food safety and public health towards a joint strategy and action plan to roll out WGS across sectors for enhanced One-Health surveillance of food-borne diseases. X EU stakeholders in public health should strengthen their collaboration with international counterparts, especially in the US, to coordinate and harmonize the application of WGS for food safety and health protection. X EFSA and ECDC should invest in the establishment of databases for molecular typing data and set up procedures to allow data reporting, access and analysis procedures for the safe and optimal use of the information to further public health. X EU stakeholders in public health should contribute to the advanced training of the people who will produce, analyse and interpret WGS data in conjunction with epidemiological data for informing risk managers in food safety and public health. X EFSA should stimulate and encourage scientific research on WGS and the implementation of the results to benefit food safety. X EFSA should initiate and coordinate the development of an EU-wide strategy for communication on food safety issues involving WGS in cooperation with ECDC. X EFSA and ECDC should instil a sense of urgency in all partners regarding implementation of WGS for food and public health safety across the EU.