Whenever you feel lonely, just remember: you're always carrying several hundred trillion friends with you. A dizzying number of microbes call the human body home, and it turns out that science knows very little about most of them. In fact, a new Stanford survey of the foreign DNA fragments circulating in the human body has found that 99 percent of microbes inside us are completely unknown to science.

The discovery was initially made by accident, as a team investigated less invasive ways to predict whether a patient's body would reject a transplanted organ. Rather than the wholly unpleasant experience of having a tissue biopsy taken, the researchers were studying whether a simple blood sample would suffice. Essentially, the idea was that if they found fragments of the organ donor's DNA circulating in a patient's blood, it was a good indication that the body was rejecting the transplant.

Along with the patient's DNA and potentially that of the organ donor, the technique gives an insight into that person's microbiome – the trillions of bacteria, viruses and other microbes that live throughout the body. Of all the non-human DNA floating around in there, the team found that a staggering 99 percent didn't match anything in existing genetic databases.

"We found the gamut," says Stephen Quake, senior author of the study. "We found things that are related to things people have seen before, we found things that are divergent, and we found things that are completely novel. I'd say it's not that baffling in some respects because the lens that people examined the microbial universe was one that was very biased."

The team then set about categorizing that pile of unknown DNA, and found that most of it belonged to a general group known as proteobacteria, which counts E. coli and Salmonella among its ranks, along with many, many others. On the virus side of things, the team found a huge amount of previously unknown members of the torque teno family, including an entirely new group that doesn't quite fit current descriptions.

"We've doubled the number of known viruses in that family through this work," says Quake. "We've now found a whole new class of human-infecting ones that are closer to the animal class than to the previously known human ones, so quite divergent on the evolutionary scale."

With so many microbes living in the human body, it's hardly surprising that science hasn't gotten around to identifying them all, and the researchers say that attention is largely focused on a few particularly interesting species. The next step, the team says, is to apply the technique to the microbiomes of other animals in order to identify viruses that could potentially jump to humans and trigger pandemics, like avian and swine flu.

The research was published in the Proceedings of the National Academy of Sciences.

Source: Stanford University