More than 2500 new viruses have been found by scanning DNA recovered from human and animal cells. The method that was used promises to identify countless more viruses.
There are untold numbers of species of virus, but only about 9000 have been characterised well enough for their genomes to be included in the definitive data base, built by a branch of the US National Institutes of Health.
A team led by Chris Buck at the National Cancer Institute in Bethesda, Maryland, had previously scanned samples of human and animal tissue for viral DNA.
Advertisement
At the time, the researchers were looking for new papillomaviruses and polyomaviruses, some of which cause cancer. Unlike many other viruses, these have genomes made of circular DNA.
鈥淚 think we鈥檙e getting close to knowing all the human-infecting species in those families,鈥� says Buck. But in the process, he says the team 鈥減ulled up a giant amount of other stuff that was not papillomavirus or polyomavirus鈥�.
This extra genetic data wasn鈥檛 studied further until graduate student Michael Tisza joined the lab. He devised a set of computer programs that could sort through it and identify new virus species.
鈥淲e鈥檝e provided a user-friendly way to sift through these junk piles,鈥� says Buck.
The analysis of the additional circular DNA data revealed 2514 new viruses. While many belong to existing families of viruses, 609 don鈥檛 resemble any known viruses.
Some of the new viruses are highly unusual. One belonged to a group called CRESS viruses, but was far bigger than any known CRESS virus. It turned out to have three copies of a gene used to make an outer shell that encapsulates the entire virus particle. The whole particle, shell and all, is called a virion.
鈥淚t seems like, by duplicating this gene a few times, the virus was able to make a bigger virion to accommodate more DNA,鈥� says Buck.
Some of the new viruses may be dangerous. For instance, the team found more than 100 anelloviruses in human blood. As yet, no anellovirus has been linked to a human disease, but conceivably some of the new ones could be harmful.
Indeed, any virus could potentially be dangerous, or evolve to be. 鈥淭he first step in finding out whether a virus is causing a disease is finding out the virus exists,鈥� says Buck.
Tisza has now developed a more advanced version of the software, which can identify other kinds of virus that lack circular DNA. In unpublished work, he has applied it to the genetic data sets of other groups.
鈥淭here鈥檚 20 petabytes of data,鈥� says Buck. 鈥淭hose data sets have already yielded 75,000 new virus species鈥�, some of which he says are 鈥渧ery exotic鈥�.
eLife
