Detective Bacteria is on the case.
In a recent paper Noah Fierer, from the University of Colorado has done some preliminary work that suggests bacteria could be used to track down criminals similar to the way we use fingerprints. Ordinarily we look for fingerprints, compare them to a database of collected fingerprints and then use that info to catch our ‘perp’. Noah’s work is based on an observation that the ‘microbiome‘ of our finger tips and hands is somewhat unique to each of us.
I don’t think we have ever mentioned it on the site, which I take the blame for as I’m the bacteriologist between us but did you know that the bacteria on and in our bodies outnumber the cells that make us up by more than 10:1? 10 bacteria for every 1 of our cells, that’s a lot of bacteria. If you look at the whole body researchers have found 19 different phyla and 205 different genera without even taking into account the hundreds of potential residents in the gut. So we have lots in terms of both numbers and types.
Anyway, Noah found that when he swabbed peoples computer mice and then scanned a database he had a collected of nearly 280 peoples ‘hand microbiomes’ he could reliably find the appropriate matches!
This is a very cool tool as fingerprints can be smudged and degrade over time but he found that even after two weeks a sample could be collected and used to get a positive result.
Whilst I think this is awesome (and don’t get me wrong, it IS cool) it’s a very long way off being used in actual crime fighting, though probably not too far away from the next episode of CSI: Arctic Circle or Law and Order and Fun Times.
The main problem, as many have pointed out, is that 280 is a pretty small number when determining the accuracy of a method like this. Also Noah only used computer mice, what about stainless steel, wood or glass? As the bacteria found on our hands is different to that found on , say, our elbows, how useful would a database of hand microbiomes be? Do the microbiomes change over time?
There is also a question of accuracy. If a fingerprint on the database only shows 90% similarity to a sample fingerprint they cannot be said to be a match, what are the thresholds for bacterial analysis?
Whilst game changing this is not, I think, with further development, this technique could become useful. Another tool in the tool belt of forensic science. It wouldn’t say much on its own but on conjunction with DNA matching and fingerprint analysis it may provide another layer of identity confirmation.