Unrestricted Versus Regulated Open Data Governance: A Bibliometric Comparison of SARS-CoV-2 Nucleotide Sequence Databases

Abstract

We analyse ongoing efforts to share genomic data about SARS-COV-2 through a comparison of the characteristics of the Global Initiative on Sharing All Influenza Data and the Covid-19 Data Portal with respect to the representativeness and governance of the research data therein. We focus on data and metadata on genetic sequences posted on the two infrastructures in the period between January 2020 and January 2023, thus capturing a period of acute response to the COVID-19 pandemic. Through a variety of data science methods, we compare the extent to which the two portals succeeded in attracting data submissions from different countries around the globe and look at the ways in which submission rates varied over time. We go on to analyse the structure and underlying architecture of the infrastructures, reviewing how they organise data access and use, the types of metadata and version tracking they provide. Finally, we explore usage patterns of each infrastructure based on publications that mention the data to understand how data reuse can facilitate forms of diversity between institutions, cities, countries, and funding groups. Our findings reveal disparities in representation between the two infrastructures and differing practices in data governance and architecture. We conclude that both infrastructures offer useful lessons, with GISAID demonstrating the importance of expanding data submissions and representation, while the COVID-19 data portal offers insights into how to enhance data usability.

Philosophy of Open Science

“…It puts a premium on fast-paced research scaffolded by high-tech infrastructurefactors which, however, are not always or necessarily marks of quality and longterm reliability (Chen et al 2019, Leonelli 2021. It can result in the exclusion from data sharing initiatives of researchers who are not based at prestigious academic institutions, which in turn reduces the diversity and range of data available online, as well as the types of expertise viewed as significant when evaluating that evidence base (Sheehan et al 2023). These are significant ethical issues in terms of who is included and recognized as a participant in research; but far from being solely an ethical problem, this form of bias has also epistemic implications, since it substantially reduces both the diversity of data sources available to biomedical researchers globally and the amount of expertiseand particularly localized, regional expertiseput to the service of data modelling and interpretation.…”

Epidemiological and Genetic Characteristics of Respiratory Viral Coinfections with Different Variants of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2)

“…Research on the genetic changes and mutations of respiratory viruses involved in such mixed infections is lacking, opening up unexplored opportunities for genetic variation. On the other hand, 6 million whole-genome sequences (WGS) of SARS-CoV-2 have been generated in EpiCoV by the Global Initiative for sharing of all influenza data (GISAID) [21]. This makes it possible to accurately calculate the evolution and transmission of SARS-CoV-2 [22,23].…”