This story is over 5 years old.

Forget Fingerprints—in the Future We'll Be Identified by Gut Bacteria

Scientists have shown for the first time that people can be identified with unique microbial “fingerprints.”
May 12, 2015, 9:00pm

Each of us comes with a unique genetic signature, or genome, containing information about we look like, what diseases we might carry, and who our dads are.

Above our own genome, however, there is another vast body of information also unique to each of us: the tens of trillions of microbes that live in and on our bodies, as well as all their genes.

Collectively known as the human microbiome, these microbes evolved with us and are intimately intertwined with many of our bodily functions. They help us ward off disease, digest our food, and might even influence our mood.


Unlike our genome, our microbiome isn't permanently fixed to us—but it might be variable enough between individuals and stable enough over time that it can be used as a distinctive identifier, similar to DNA.

Compared to microbes in the skin, mouth, and vagina, gut microbes were the most stable over time

In a paper published yesterday in PNAS, scientists reported that they were able to identify unique microbial "fingerprints" for more than one hundred participants of the Human Microbiome Project, a research initiative funded by the National Institutes of Health.

Furthermore, the researchers observed that some of these fingerprints stayed relatively stable over the span of a year. Over 80 percent of the time, they could correctly identify an individual by matching gut samples taken from that person a month to a year apart.

These findings could add another layer to personalized medicine, the emerging practice of tailoring medical treatments to individual patients based on their genetic profiles, according to Eric Franzosa, a research associate at the Harvard School of Public Health and lead author of the PNAS paper.

"Hopefully one day this kind of personalized information could be used to help people make decisions about their own health," said Franzosa. "For example, if your microbiome is different from mine, it might make sense for you to eat differently or take different medications."


Franzosa gave the example of digoxin, a drug used to treat heart failure. Researchers have found that some people have gut microbes that digest and deactivate the drug, while others don't. Knowing which microbes a patient has could help doctors decide which medicine to prescribe, said Franzosa.

To identify personalized microbial signatures within their study population, the researchers created an algorithm based on what's called a "hitting set" approach. According to Franzosa, the approach works by trying to assign an identifier to each person that is present in that individual but absent in everyone else.

"Say you give me a collection of a bunch of people's lists of favorite movies," said Franzosa. "How do you find a list of movies where each person likes one of those movies—or some 'set' that 'hits' each person's preference?"

Franzosa and his colleagues set up their algorithm to search through the data of Human Microbiome Project participants and find microbial genetic patterns unique to each person. Overall, they created microbial fingerprints for 120 people.

Compared to microbes in the skin, mouth, and vagina, gut microbes were the most stable over time, according to the researchers. Around 86 percent of the time, the scientists were able to uniquely identify an individual based on stool samples taken 30 to 300 days after the original stool sample from which the microbial fingerprint was derived. With bacteria in the skin, mouth, and vagina, later fingerprints matched the original ones about 30 percent of the time.


In their paper, the authors acknowledged that their findings could have ethical implications on biomedical research involving the human microbiome. Chiefly, study participants should be aware that their microbial genetic data could be used to identify who they are and possibly convey information about their health, diet, age, and geography.

"It's important for researchers and participants to know the possibility exists," said Franzosa. He thinks a good rule of thumb is for scientists to follow the same precautions that they would for research involving human genetic information. Informed consent is one important aspect of that, he said.

The authors of this study are the first to actually show that the microbiome can be used to identify people, according to Dan Knights, a microbiome researcher and professor of computer science and engineering at the University of Minnesota. "They have shown that an individual can be identified by their gut bacteria using a sample taken in the past," he said. "Previous studies have only demonstrated this in theory."

Microbial fingerprints could prove useful in forensics, said Knights, particularly in "choosing a suspect from a small lineup for a crime where someone defecated on the crime scene"—something that is not actually that uncommon, he added. However, suspects could possibly "cheat" the test by taking antibiotics that alter the bacteria in their gut, Knights speculated.


That said, there are limits to using the microbiome as an identifier, according to Knights. He pointed to the fact that people tend to lose some of their unique fingerprint over time, and that microbial fingerprinting breaks down when trying to identify someone from a population larger than a few hundred people.

The unreliability of microbial fingerprints limits potential forensic applications of the microbiome, according to David Rasko, a professor of microbiology at the University of Maryland. "I think many people will automatically think of using this for the identification of individuals in crimes or other legal identification," he said. "However, I think that this is still a long way off."

Researchers first need to better understand the dynamics of the human microbiome before jumping to use in forensics, said Bruce Roe, a professor of biochemistry at the University of Oklahoma. "The results in identifying individuals are not of the accuracy level even close to present-day forensic methods," he said.

DNA will probably continue to win over the microbiome because the human genome is much more unique

How our individual microbial signatures become so distinct is still an open question, according to Franzosa. Many factors seem to play a role, including genetic background, medications, diet, disease, and environmental exposure, he said.

For Franzosa, the most interesting application of this research isn't in traditional forensics. For something like identifying biological samples at the scene of a crime, DNA will probably continue to win over the microbiome because the human genome is much more unique, he said.

But for something like revealing the places a person has recently visited, the microbiome could come in handy, said Franzosa. Researchers can find distinctive microbial signatures for various places, including natural environments like ponds and built environments like buildings and trains. In forensics, that information could help reveal where a person had recently been, Franzosa speculated. "It's an interesting forensic application in which the microbiome could even beat the human genome," he said.

Moving forward, Franzosa wants to find other ways to harness the remarkable dynamism of the microbiome. "I'm interested in exploring things that are unique about the microbiome—like the fact that it can change during life. What sorts of information can you learn from that?"