Unveiling Origin Pitfalls in Metagenomic Surveillance for Emerging Infectious Diseases: Parvoviruses as a Model

ABSTRACT Emerging infectious diseases caused by zoonotic pathogens pose significant threats to human health. Metagenomics‐based surveillance is essential for emerging pathogen identification. During clinical surveillance, we identified potential parvovirus‐related spillover events from birds and mammals. However, labor‐intensive investigations revealed that silica membranes in nucleic acid extraction kits harbor parvoviruses across all three subfamilies. The silica membrane virome uncovered misattributions involving at least 13 viral families, with parvoviruses as the predominant component, complicating global genomic surveillance. Here, we used parvoviruses as a model, utilizing ChatGPT‐4o to curate the Panoramic Virus Discovery Data Chain, which documented the comprehensive virus discovery process. This revealed substantial uncertainties in parvovirus‐host relationships and identified shared parvoviral species between humans and nonhuman primates. To address these expanding challenges, we developed the Parvovirus Database, enhancing global parvovirus–host–disease associations. This study underscores the methodological pitfalls in metagenomics‐based surveillance and establishes a framework for accurate associations among viruses, hosts, and diseases.