The mouse gut virome: 109,000 viral genomes and the ageing signals bacteria miss
The gut microbiome has been one of the most active frontiers in biomedicine over the past two decades, but the research has had a blind spot. Most studies focus on bacteria, the hundreds of species that live in the intestines and influence everything from digestion to immunity. The viruses that infect those bacteria, known collectively as the gut virome, have been largely ignored, mainly because they are harder to catalog.
A new study in Nature Communications changes that, at least for mice. Researchers led by Insuk Lee at Yonsei University in South Korea have produced the Mouse Reference Gut Virome (MRGV), a comprehensive catalog of 109,778 viral genomes representing 28,824 species. The resource expands known mouse gut viral diversity by about 67.6 percent and, perhaps more importantly, reveals that the virome carries signals of aging that the bacterial microbiome cannot detect.
Why the virome matters
Viruses that infect bacteria, bacteriophages, are not passive passengers in the gut. They shape bacterial community structure by killing specific strains, transferring genes between bacteria, and influencing bacterial metabolism. Through these effects, they can alter the host’s immune system, inflammatory状态, and even neurological function. But studying them has been difficult because phages are genetically diverse, rarely culturable, and poorly represented in existing sequence databases.
The MRGV changes that by providing a reference-grade resource. The researchers assembled viral genomes from fecal samples using a technique called viral genome binning, which groups short DNA sequences into complete or near-complete genomes by looking at co-occurrence patterns across samples. This is a significant methodological advance: binning accounted for about 36 percent of the cataloged genomes and improved average completeness by 60 percent compared with traditional contig-level assembly.
A massive protein annotation effort
The scale of the resource is striking. The team predicted 8.2 million viral proteins from the newly assembled genomes. Using structure-based annotation methods, predicting protein function from 3D structure rather than just sequence similarity, they assigned functions to nearly half of them. This is important because most viral proteins have no sequence similarity to anything in existing databases; structure-based approaches can catch functional relationships that sequence-only methods miss.
Functionally, the virome is not a random collection. The team showed that about 88 percent of viral genomes could be linked to a specific host bacterium, and they validated diverse mouse-associated crAss-like phage lineages, a group known from human guts, with lineage-specific host range and replication strategies.
Aging in the virome
The headline finding is that the mouse gut virome contains features that correlate with aging, and these features are distinct from those captured by bacterial microbiome analysis. In other words, measuring only the bacterial members of the gut ecosystem misses an entire dimension of age-related change.
Most of the aging-associated markers came from the binned genomes, the ones assembled using the more sophisticated binning approach, rather than from simple contig-level assemblies. This underscores the importance of the methodological work: without viral genome binning, the aging signal would have been substantially weaker.
The specific changes include shifts in phage population structure, with certain viral lineages expanding or contracting as mice age, and corresponding changes in the functional capabilities of the virome. These patterns may reflect age-related changes in the bacterial community or in the host immune system, or both.
A resource for the field
The MRGV is now publicly available and will serve as a baseline for future studies of the mouse gut virome in health and disease. The researchers have made all 109,778 genomes, the 8.2 million protein annotations, and the host linkage data freely accessible.
The study also highlights a gap. The mouse and human gut viromes show “pronounced taxonomic and functional divergence,” meaning findings from mice cannot be directly extrapolated to humans. A similar effort for the human gut virome, at this level of completeness, would be the logical next step.
For now, the message is clear: the virome is not noise. It carries information about host biology that the bacterial microbiome does not, and ignoring it means leaving signals on the table.
Reference: Kim et al., “A genomic catalog of the mouse gut virome reveals features associated with ageing,” Nature Communications (2026). DOI: 10.1038/s41467-026-75836-6.