Host–virus association databases as tools for understanding viral spillover at varying scales

Large host-virus association databases are increasingly used to explore broad questions in disease ecology, particularly around host range, pathogen diversity, and the potential for spillover. While these databases have been instrumental in large-scale synthesis of host-pathogen biogeography and zoonotic risk, their potential role in addressing fine-scale questions about pathogen prevalence, maintenance, and transmission dynamics remains underexplored. In this study, we build on previous efforts to assess how different types of data, including both entries in databases and the original studies they draw from, can support targeted research on zoonotic spillover. We selected two zoonotic diseases, Ebola virus disease and Lassa fever, which are characterised by recurrent spillover events and outbreaks in sub-Saharan Africa. We searched the VIRION database for entries corresponding to the respective viral taxa, the genus Orthoebolavirus and the species Mammarenavirus lassaense, and used these entries as case studies. We evaluated the extent to which databases capture crucial contextual metadata, such as spatial and temporal resolution, negative results, and measures of viral load. Guided by a conceptual framework of factors that lead to spillover, we demonstrate that while host-virus databases are valuable for addressing high-level patterns, fine-scale investigations of spillover require specific studies with detailed epidemiological data. Our study adds to a growing body of literature offering practical recommendations for database users and managers and highlights how these tools can be used as starting points in spillover research.