Ancestry Informative Markers in the Skin Microbiome for Human Identification

Over the past decade microbial forensics has shifted from solely biosecurity applications to include human identification (HID). Studies have shown that the human skin microbiome could potentially serve as an adjunct to traditional DNA evidence, especially when the human DNA is low quantity. The 16S rRNA marker has been the focus of most forensic microbiome studies for HID. This marker, however, has fallen short of consistent and accurate identification with large sample sizes. While shotgun sequencing provides, in theory, greater breadth and depth of information about the skin microbiome, specifically subspecies and strain level differences, the read coverage and depth are stochastic, reducing the accuracy for HID.  Recently, a custom targeted panel for the human skin microbiome has been developed. The panel leverages loci specific to key microbes that are highly abundant and stable on human skin. Leveraging genetic variants in stable microorganisms may be one promising approach to microbial HID, however how such variants should be selected is an open question. Ancestry informative markers (AIMS) provide one potentially powerful approach. In human populations, AIMs leverage single nucleotide polymorphisms (SNPs) with allele frequencies that differ substantially between (but not within) populations to predict bio-ancestry. Extending this concept to the person as a population of microbes, AIMs selected based on F_ST (Wright’s index of fixation) can serve as a promising approach to identify the human host. Skin swabs from the non-dominant hand of 51 individuals were collected in triplicate and were analyzed for HID purposes. High F_ST SNPs were selected using three different methods to determine if the number of taxa and/or SNPs had an impact on HID accuracies. The SNPs were then input into a support vector machine (SVM) to classify unknown samples to the individual from which they most resembled. Accuracy of classification ranged from 88% – 95%, suggesting that AIMs in targeted microorganisms can improve accuracy of HID. The data support that AIMs from the microbiome may also be a powerful tool for HID.