PVC Release V0.1

This table contains information on the viruses currently included in the most recent PVC release. Please see the other files in this project to download sequences and additional metadata.Abstract<br><br>Identifying viruses in compositional bulk sequencing data, like metagenomics or metatranscriptomics, is of increasing interest to the scientific community. However, small, repetitive, and poorly characterized genomes haunt viral reference databases, so accurately estimating the viral composition of a community is difficult. Here, we present both 1) the Pan Viral Compendium (PVC), a quality-controlled database generated from known viral life alongside 2) an aligner, Xtree, optimized for clade-specific, high sensitivity and specificity viral genome alignment. The PVC is a dereplicated set of viral genomes from nine estabilished databases, representing 2,851,990 viruses when dereplicated at 99% identity. We provide taxonomic and host species annotations, quality control metrics for every genome, and indexed alignment databases for use with Xtree. We additionally benchmarked Xtree, identifying its overall accuracy in viral identification in complex communities across taxonomic ranks and within specific clades, identifying parameters (e.g, specific databases, coverage thresholds) for quantifying precision viral abundances at even the species level. In total, the PVC and its companion aligner will enable strain-resolved viral metagenomics and metatranscriptomics that can be easily optimized to a given user's specific needs.