https://phys.org/news/2018-06-burglary-microbiome.html

The researchers collected samples from residents (nostril and hand) and various surfaces from their respective homes in Chicago, IL and Fort Lauderdale, FL prior to and following mock burglaries by non-residents. They identified unique bacterial assemblages for each individual, and generated models to discern the accuracy of predicting a non-resident interacting with a given home.

Humans emit approximately 36 million microbial cells per hour into their immediate environment, providing the opportunity to trace microbial signatures back to their originating source. The potential of tracking microbial exchange between an individual and a built environment surface has been well demonstrated by matching the microbial signature on individual participants' finger tips to keys on a computer keyboard, as well as personal devices.