Comparison of metagenomic protocols for virome data generation from environmental matrices and stool samples: insights into viral diversity and fecal contamination indicators

Anthropogenic factors play a critical role in the spread and emergence of viruses, and the investigation of reliable markers of human-derived pollution is key to properly monitoring environmental quality. This study examines viral diversity in 60 human stool samples, 12 surface water samples, and 12 groundwater samples, all collected within the same geographic scenario, namely the Rio dos Sinos Watershed in Southern Brazil, employing four distinct sample preparation protocols for metagenomic. The results demonstrate that viral diversity and abundance depend on the preparation protocol used. A total of 139 genera across viral 55 families were identified, encompassing a wide range of genome types (dsDNA, ssDNA, ssRNA, dsRNA) and hosts (bacteria, plants, humans, and other mammals). Among the key findings are the characterization of six new CrAssphage genomes, which exhibit strong potential as indicators of human fecal contamination, and the identification of Microviridae genomes with significant genetic variability, suggesting the existence of new subfamilies. Additionally, three new Pepper mild mottle virus (PMMoV) genomes were classified into pathotypes sensitive to the L3 resistance gene, offering insights for managing agricultural diseases. As the first non-targeted metagenomic study of viral diversity in this region, this research provides a foundation for understanding virome composition and its broader implications for public health, agriculture, and environmental sustainability. The study emphasizes the need for protocol optimization and further exploration of viral communities to enhance surveillance and mitigation strategies.