Comparative genomics of bat endogenous retroviruses: abundance and coding integrity are decoupled, and de novo detection can counterfeit a lineage-specific expansion

Deep inside the DNA of every mammal, from mice to humans, lies a hidden archive of ancient viral infections. These are not active viruses waiting to strike, but rather the fossilized remains of retroviruses that successfully invaded the reproductive cells of our ancestors millions of years ago. Once a virus inserts its genetic code into a host's germ line, that code becomes a permanent part of the family tree, passed down from generation to generation. Over time, most of these viral invaders lose their ability to function, accumulating mutations and breaking apart until they are just scattered fragments of genetic code. Scientists call these endogenous retroviruses. They act as a record of the battles between hosts and viruses in deep time, and because different groups of animals have faced different viral histories, the amount and condition of these viral fossils vary wildly across the tree of life.

Bats are a particularly fascinating group to study in this context. As the second most diverse order of mammals, with over 1,400 species, they are also famous for their ability to carry a vast array of viruses without getting sick. This unique tolerance has made them a subject of intense interest for researchers trying to understand how viruses evolve and how hosts survive them. However, when scientists began comparing the viral fossils inside the genomes of different bat families, they ran into a problem. The data they were looking at was often unreliable. The genomes they used were sometimes mislabeled, containing the DNA of the wrong species, or the methods used to count the viral fossils were so sensitive to the way the DNA was assembled that they created false signals of recent viral activity. To get a true picture of bat viral history, researchers first had to build a framework that could separate real biological facts from computer-generated errors.

A team of researchers set out to create this trustworthy framework by analyzing the genomes of 29 different bat species, representing 11 distinct families. Their first task was to verify that every genome they used actually belonged to the bat species it was claimed to be. They cross-referenced the official labels on the data with three independent lines of evidence: the metadata provided by the database, the naming conventions used by major bat genome projects, and the genetic sequence of a specific gene found in the mitochondria, the energy centers of the cell. This rigorous check was necessary because a single mislabeled genome could completely distort the comparison, making it look like a virus jumped between unrelated species when it had not. After this screening, they removed three assemblies that were misidentified and excluded five others that were too fragmented to analyze reliably, leaving them with a clean, verified panel of 29 genomes.

With a solid foundation of verified data, the team turned to the viral fossils themselves. They used a computer pipeline to scan these genomes for long terminal repeats, the signature structures that mark the beginning and end of a retroviral insertion. They found that the amount of viral DNA varied dramatically between families. In some groups, like the fruit bats known as Pteropodidae, viral fossils made up only about 1.7% of the genome. In others, such as the mouse-tailed bats (Rhinopomatidae), the figure soared to nearly 19%. This was an eleven-fold difference. Crucially, the researchers proved that this variation was a real biological trait and not an artifact of how well the DNA was pieced together. They showed that the low amount of viral DNA in fruit bats was consistent across five different, independently assembled genomes and could not be explained by the quality or completeness of the DNA data.

The study also revealed a surprising disconnect between how much viral DNA a bat has and how well-preserved it is. One might assume that families with the most viral DNA would also have the most intact, functional-looking viruses. Instead, the researchers found that abundance and integrity were largely independent. A family could be rich in viral sequence but mostly filled with broken, degraded fragments, while another with less total sequence might retain a higher proportion of complete, structurally sound viral fossils. This suggests that the history of viral infection and the history of viral decay are two separate stories. For example, fruit bats, despite having very little viral DNA overall, stood out because the few complete viruses they did retain were unusually likely to carry a specific gene called env, which codes for the viral envelope. This gene is often the key that allows a virus to enter a cell, and its presence in these ancient fossils hints that these specific viral remnants may have been preserved for a reason, perhaps because the host animal found a use for them.

Perhaps the most significant finding of the study was a warning about how scientists interpret these genetic landscapes. In one specific genome, the initial computer scan suggested a massive, recent explosion of viral activity, showing a huge number of young viral fossils that seemed to have inserted themselves very recently in evolutionary time. This looked like a dramatic biological event: a lineage-specific burst of infection. However, when the researchers looked closer, they realized this was a trick of the light. The genome in question was assembled at a lower level of detail than the others, meaning the DNA was broken into many small, disconnected pieces. This fragmentation caused the computer to count the same viral fragment multiple times and to make the viral copies look younger than they really were. When the researchers checked the same genome using a different method that did not rely on the fragmented assembly, the "explosion" vanished. The bat actually had a normal, low level of viral DNA, indistinguishable from its close relatives.

This discovery serves as a vital safeguard for future research. It demonstrates that raw computer counts of viral fossils can be easily fooled by the way DNA is assembled, creating the illusion of a recent biological event where none exists. The researchers concluded that before accepting any claim of a unique, lineage-specific expansion of viruses, scientists must confirm the finding with evidence that does not depend on the assembly structure, such as direct comparisons of specific gene sequences. By applying these strict controls, the team provided the first reliable baseline for understanding the viral history of bats. They showed that while fruit bats have indeed suppressed their viral load, the story of viral evolution in bats is complex, with abundance and preservation telling different chapters of the same history, and that the most dramatic signals in the data are often the ones that need the most careful scrutiny.