IMG/VR v4: an expanded database of uncultivated virus genomes within a framework of extensive functional, taxonomic, and ecological metadata

Viruses are widely recognized as critical members ofall microbiomes. Metagenomics enables large-scaleexploration of the global virosphere, progressivelyrevealing the extensive genomic diversity of viruseson Earth and highlighting the myriad of ways bywhich viruses impact biological processes. IMG/VRprovides access to the largest collection of viral se-quences obtained from (meta)genomes, along withfunctional annotation and rich metadata. A web inter-face enables users to efficiently browse and searchviruses based on genome features and/or sequencesimilarity. Here, we present the fourth version ofIMG/VR, composed of > 15 million virus genomesand genome fragments, a asymptotic to 6-fold increase in sizecompared to the previous version. These clusteredinto 8.7 million viral operational taxonomic units, in-cluding 231 408 with at least one high-quality repre-sentative. Viral sequences in IMG/VR are now sys-tematically identified from genomes, metagenomes,and metatranscriptomes using a new detection ap-proach (geNomad), and IMG standard annotation arecomplemented with genome quality estimation us-ing CheckV, taxonomic classification reflecting thelatest taxonomic standards, and microbial host tax-onomy prediction. IMG/VR v4 is available athttps://img.jgi.doe.gov/vr, and the underlying data areavailable to download athttps://genome.jgi.doe.gov/portal/IMGVR.