Gene networks provide a high-resolution view of bacteriophage ecology
Bacteriophages are the most abundant and diverse biological entities on the planet, and new phage genomes are being discovered at a rapid pace from metagenomes. As more novel, uncultured phage genomes are published, new tools are needed for placing these genomes in an ecological and evolutionary context. Phages are difficult to study with phylogenetic methods, because they exchange genes regularly, and no single gene is conserved across all phages. Instead, genome-level networks have been used to group similar viruses into clusters for taxonomy. Here, we show that gene-level networks provide a high-resolution view of phage genetic diversity and offer a novel perspective on virus ecology. To that end, we developed a method that identifies informative associations between a phages annotated host and clusters of genes in the network. Given these associations, we were able to predict a phages host with 86% accuracy at the genus level, while also identifying genes that underlie these virus-host interactions. This approach, thus, provides one of the most accurate means of host prediction while also pointing to directions for future empirical work.