What will the next big zoonotic virus be? Most emerging infectious diseases are zoonoses, and many, if not most, of them are caused by viruses, but right now we have little idea what viruses are out there in livestock and wildlife, let alone which one might be next in line to cause a human pandemic. In their study in mBio this week, a group from Columbia University, the EcoHealth Alliance, and other institutions undertake an exhaustive search for viruses in one species, then use those results to extrapolate the total viral diversity in all mammalian species. They estimate that all the mammals in the world harbor a total of only 320,000 viruses and that it would cost only $6.3 billion to discover them all - a pittance compared with beating back a global pandemic like SARS.
"I was surprised," says co-author Peter Daszak. The number of mammalian viruses, says Daszak, "is not tens of millions, it's hundreds of thousands. It's a number we can deal with."
The total estimates of viral diversity and the money it will take for discovery started with a study of a single species: the Indian flying fox. With 1,897 samples of urine (n = 926), throat swabs (n = 806), feces (n = 78), and roost urine (n=97) from apparently healthy Pteropus giganteus bats throughout Bangladesh, they used PCR with a set of degenerate viral family-level primers. This search turned up only 55 unique viruses, each as different from one another as the two most closely related viruses in their respective families.
At this point, we get into some serious modeling and statistics. Using rarefaction curves and nonparametric estimators, the authors calculated that the maximum number of viruses harbored by P. giganteus must be around 58. (Details here.)
The 320,000 figure was easier to calculate: they used 58 as the average number of unique viruses in each of the 5,486 described mammalian species and multiplied. Extrapolating from the costs involved their own study of P. giganteus, they estimate that the cost of sampling and viral discover for all mammalian viruses would be around $6.3 billion. Because of the diminishing returns of an exhaustive search for every last virus, identifying 85% of these viruses would come at a much smaller cost of $1.4 billion.
But Daszak argues that $6.3 is a small price to pay for the potential to predict the next zoonotic pandemic. "If you look at every pandemic in past 50 years, almost every one is a virus from wildlife," says Daszak. "We spend hundreds of billions when a pandemic emerges. A one-off cost of less than $10 billion and we can potentially avoid the risk to future health."
Daszak acknowledges that many assumptions went into these calculations. The PCR primers only detect nine virus families, there may well be considerable overlap between viral hosts, and some species may be host to more unique viruses than others, to mention a few. But these assumptions may over- or under-estimate the real number, Daszak says, so he's comfortable that they're in the right ballpark.
Daszak says he and his colleagues are busy applying their viral search methods to new species in order to get a better handle on the average number of viruses per mammal. "We're going for primates first," says Daszak. He's hopeful that viral prospecting will lead the way to better public health.
"In 50 years, predicting pandemics will be part of our standard operating procedure for dealing with emerging diseases," says Daszak.