Answers are grounded in this paper and cite their sources.
Origin and first weeks The Virus Outbreak Data Network (VODAN) Implementation Network (IN) Footnote 1 was conceived to kick-start a ‘community of communities’ that could design and rapidly build a truly international and interoperable, distributed data network infrastructure that supports evidence-based responses to the viral outbreak. The IN has a longer-term goal to reuse the resulting data and service infrastructure, also for future outbreaks. As a GO FAIR IN, VODAN will restrict itself to projects that are directly associated with FAIR data [ 1 ] and services relevant to COVID-19. This also means that VODAN will not concern itself with projects purely aimed at studying, for instance, transmission and severity genetics, experimental drug development, vaccine development or actual clinical interventions currently executed to control the epidemic. Only insofar as such activities are in need of access to, and processing of, FAIR data, VODAN partners can add FAIR data and services value. Early on in the process, CODATA Footnote 2 , the committee of the International Science Council Footnote 3 , the Research Data Alliance (RDA) Footnote 4 and the World Data system Footnote 5 now also operating in joint actions under the name ‘Data Together’ Footnote 6 , reviewed the initial actions of VODAN and issued a joint statement Footnote 7 of support. At the request of the European Commission, RDA established a Working Group Footnote 8 , which is also co-supported by the Data Together organisations. IN–IN collaboration: As a typical, open, GO FAIR Implementation Network, VODAN is open to all parties, public or private, who commit to follow the FAIR guiding principles and who explicitly wish to prevent vendor lock-in (see GO FAIR rules of engagement: https://www.go-fair.org/resources/rules-of-engagement/ ). The VODAN partners rely on the communities already established in other GO FAIR INs, including the Personal Health Train Footnote 9 , IN-Africa Footnote 10 , the Data Stewardship Competence Centers Footnote 11 , C2CAMP Footnote 12 (for core infrastructure), Chemistry Footnote 13 , CO-OPERAS Footnote 14 (for SSH data), DISCOVERY Footnote 15 (search, visualisation) and FAIR StRePo Footnote 16 (standards and policies). Since its approval as an IN on March 13, the number of partners in VODAN increased much faster than in regular INs and on April 10, the network already approached 150 partners (Fig. 1 ) This indicates a perceived need for community based networking. Fig. 1: Accumulated number of VODAN IN members. The rapid growth of partners in the first weeks of the virus Outbreak Data Network. Full size image The current VODAN partners include both public and private parties who united around the notion that FAIR data will be needed in many aspects of an adequate response to the current, as well as to future infectious disease outbreaks. In VODAN, we clearly distinguish between ‘established knowledge’ (EK) based on the published literature, databases and public datasets, and what we call ‘Real World Observations’ (RWO), made on actual situations in real time. The main challenge of VODAN is on the latter class of data, as many other public as well as private organisations are already dealing with FAIR-enabled public databases and applications (such as for instance ELIXIR, BBMRI, EBI, NLM, Euretos). Those organisations that work on EK bases and biomedical research results are obviously candidate ‘data stations’ in the Personal Health Train sense and these are welcome to the VODAN IN and to format their assets also according to the FAIR principles. Rationale: Why is access to the necessary RWO, high quality, data on the current epidemic so difficult? First, there is the established ‘GDPR’ issue of citizen/patient privacy and this means that the RWO cannot just be ‘open’. VODAN will also demonstrate the difference between ‘FAIR’ and ‘open’ by combining restricted data with opne data [ 2 ]. Second, this epidemic is particularly highly politicised, and there is little chance that countries (or even institutions) will ‘share their RWO data’ in the classical way, i.e. by ‘sending’ them to a central data warehouse, not even at WHO. A third issue to address is that, like in the earlier HIV epidemic, there will be so much money forthcoming for research and experimental interventions that there will be an enormous variety of overlapping, and even redundant projects and consequently a wave of data, of which a significant portion will be of questionable quality. This means that we need a strong emphasis on distributed analytics as well as on vetting and annotation of data, for which we need (expert) distributed annotation in near-real time. This can only be done with machine-assistance and through (expert) community-based massive annotation. These considerations are not unique to the current COVID-19 epidemic and the mitigation of these problems is at the core of high-quality, reproducible and peer reviewed science leading to evidence-based interventions. In other words, a modern ‘epidemic-ready data and services infrastructure’ would have to address all these issues very effectively. It was from this perspective that we came up with the boundary conditions for a ‘COVID-19 data and services infrastructure’ in the common Data Together statement.
No takes yet. Share an insight, caveat, or question.
Barend Mons (2020) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: