SRA

Microbial communities play an important role in organismal and ecosystem health. While high throughput metabarcoding has revolutionized the study of bacterial communities, generating comparable viral communities has proven elusive, particularly in wildlife samples where the diversity of viruses present and limited quantities of viral nucleic acid present distinctive challenges. Metagenomic sequencing is a promising solution for studying viral communities, but the lack of standardized methods currently precludes comparisons across host taxa or localities. Here we developed an untargeted shotgun metagenomic sequencing protocol to generate comparable viral communities from non-invasively collected fecal and oropharyngeal swabs. Using samples from common vampire bats (Desmodus rotundus), a key species for virus transmission to humans and domestic animals, we tested how different storage media, nucleic acid extraction procedures and enrichment steps affect viral community detection. Based on finding viral contamination in fetal bovine serum, we recommend storing swabs in RNALater or another non-biological medium. We recommend extracting nucleic acid directly from swabs rather than from supernatant or pelleted material, which had undetectable levels of viral RNA. Using a low-input RNA library preparation protocol, we established that ribosomal RNA depletion and light DNAse treatment reduce host and bacterial nucleic acid and improve virus detection. Finally, applying our approach to twelve pooled samples from seven localities in Peru, we showed that detected viral communities saturated at the attained sequencing depth, allowing unbiased comparisons of viral community composition. Future studies using the methods outlined here will elucidate the determinants of viral communities across host species, environments and time.