The microbial communities living on and in human bodies—the human microbiome—contain unique ‘fingerprints’ that can be used to identify people, a study finds. Researchers from Harvard T.H. Chan School of Public Health have for the first time revealed that personal microbes have a strong enough variation to contain distinguishing features, which could be used to identify hundreds of individuals over time.
Using publicly available microbiome data from the Human Microbiome Project (HMP), researchers examined the microbes present in and on a number of body sites including stool, saliva and skin from 242 individuals. By using an adapted classical computer science algorithm, researchers were able to create ‘codes’ from distinguishing and stable sequence features from initial microbiome samples.
These codes were then compared to microbiome samples collected during follow-up visits from the same individuals and from independent groups of individuals. The results showed a strong variation in the microbial community structure between individuals.
“Microbiome ‘fingerprints’ provide a view of the personalization of the human microbiome,” says lead author Eric Franzosa, a research fellow in the Department of Biostatistics at Harvard Chan. “Similar to the human genome, there are many aspects of the microbiome that are conserved between individuals, but the differences that exist are sufficient to distinguish us."
Forensic genetics, a now decades-old field, has shown how small differences in the human genome are used to identify individuals based on genetic material. Franzosa and his research team suggest that the same methods could be applied to genetic information from the microbiome in cases where human DNA is unavailable or badly degraded. While the human genome is roughly constant throughout life, Franzosa says the microbiome is capable of picking up features from the environment that may aid in identifying locations an individual has visited.
“Outside of forensic applications, individual-specific features of the human genome are also considered promising targets for personalized medicine. In the future, personalized features of the human microbiome may be similarly targeted to promote human health or to treat diseases that are linked to perturbations of the microbiome,” Franzosa tells IFLScience.
The study, published in the journal PNAS, shows that a large number of these ‘microbial fingerprints’ remain stable over a year-long sampling period, which raises ethical concerns for people enrolled in human microbiome research projects. Gut microbes were particularly stable as 80% of individuals could still be uniquely identified for up to a year after the initial sampling period.
“If an individual contributed microbiome samples to two databases, it would be possible to link that individual between the databases, along with potentially sensitive metadata or sensitive information, such as the presence of a sexually transmitted infection, inferred from the microbiome data itself,” Franzosa says.
The same sort of linking would be possible if someone collected a new microbiome sample, which could be potentially connected to an individual in an online database. The likelihood of microbiome data being abused in this manner is probably low, Franzosa says, but it is important for researchers and study participants to know that the theoretical possibility exists.
Franzosa and his research team hope to explore the microbes that exist in environments outside the human body, such as bodies of water or human-built structures. As microbiome data for such environments become readily available, Franzosa and his team are “excited” to apply their methods to those data to “see if non-human environments also possess unique ‘microbiome fingerprints’ and, if so, to explore their potential applications."
2