Wiping fingerprints and cleaning up stray hairs may not be enough to remove evidence that you've been at a crime scene.
According to microbial ecologist Rob Knight, speaking at TED 2014 in Vancouver on Wednesday, the bacteria on your hands leave a "print" on a computer mouse that can be identified with up to 95% certainty— and potentially used by forensic investigators.
Two humans have 99.9% similar DNA, but could have wildly different bacteria living on their skin. The species of bacteria found on their palms would only be about 13% similar, according to a study that Knight co-authored in 2010.
The study tested whether or not skin bacteria could be easily recovered from surfaces like keyboards and computer mice — and if that bacteria would identify the individuals who had left it behind.
Back in 2010, the new technique was 70-90% accurate, but Knight's talk suggests that the rate of accuracy is rapidly improving.
How your bacteria to identify you
When blood, hair, semen, saliva, or tissue is found at a crime scene, investigators can use it to examine the DNA of the human who left it behind. A good DNA sample is considered incredibly accurate evidence, but samples can be damaged, misleading, or improperly handled. In some cases, it's impossible to find DNA at a crime scene at all.
But if someone leaves behind traces of their skin's unique bacterial colonies when they touch things, the DNA from those traces could potentially be matched back to the individual it came from. Such a match could serve was a confirmation of human DNA evidence or provide new leads in cases where no DNA exists.
To test this theory, Knight and his collaborators compared the bacteria living on three keyboards to the bacteria living on their owners' fingers. They were different enough to show who had used which keyboard.
Then they checked other public keyboards, to see if the original three volunteers in the study could be linked to any of those. They couldn't be.
The unique bacterial traces that researchers had swabbed from the keyboards were still there, even after sitting in an average temperature room for two weeks — suggesting that the sampling technique might be usable in slow-moving forensic investigations. Or crime scenes stumbled on after the fact.
In a follow up study the researchers tested the hands and computer mice of nine additional people and let the computer mice sit untouched for more than 12 hours before collecting bacterial samples.
In every case, the bacteria on the mouse matched up with the bacteria on its owner's hand, significantly more so than with samples from 270 hands that had never touched the mouse.
Proceed with caution
The authors of the study caution that forensic analysis requires considerable testing and refinement before it can be regularly used in criminal investigations. And right now, a good DNA sample is still more accurate.
But they said that their results show that analyzing bacteria could provide independent confirmation of other forensic findings like fingerprints or human DNA, and could provide evidence in cases where those don't exist. In the future, a refined version of microbe-based forensics could potentially provide a more "personally identifying" picture of an individual than other commonly accepted techniques.
Plus, this technology has already been tested where it counts, as Knight explained in his TED talk. "It was featured on CSI Miami, so you really know it's true," he said.
SEE ALSO: The 40 Science Experts Will Completely Revamp Your Social Media Feed