Skip to content

Study Tests Microbial Traces for Forensic Identification

New forensic research tests microbial DNA from soil and touched surfaces, revealing promising location attribution but variable donor identification results.

By 2 min read

Researchers in the United States have investigated whether microorganisms recovered from soil and touched surfaces could help forensic scientists determine where evidence originated or identify possible contributors.

The study, published online on 16 November 2026 in the Journal of Forensic Sciences, was conducted by Emily Cantrell and colleagues at the University of Tennessee, Knoxville.

Microorganisms are present throughout the environment and on human skin. Their communities can vary between locations and individuals, making them potential sources of additional forensic information.

The researchers examined two types of simulated trace evidence: soil transferred from four source locations and microbial material deposited through human touch.

They analyzed microbial communities using 16S ribosomal RNA gene sequencing, a technique that helps characterize bacteria through genetic markers. The resulting data were evaluated using an extreme gradient boosting (XGBoost) machine-learning classifier.

For soil evidence, the researchers reported strong location attribution performance. Classification accuracy exceeded 84% in the evaluated sink samples and reached 97.9% under some testing conditions.

These findings suggest that differences in microbial communities could help associate soil traces with their geographical sources, at least within the controlled experimental setting.

However, the results for human touch evidence were less consistent.

In one testing configuration, the model achieved 86% top-three accuracy when associating microbial traces on polyvinyl chloride surfaces with donors. Performance declined to 68% when the transferred samples underwent treatment to remove extracellular DNA.

Extracellular DNA, sometimes called relic DNA, is genetic material found outside intact living cells. Researchers investigated whether removing this material would improve the interpretation of microbial evidence.

Their findings showed that the effect depended on the evidence type. Removing relic DNA did not consistently improve classification and could reduce the information available from touch samples.

The research is significant because it provides experimental performance benchmarks for microbial trace evidence across different sample types and preparation conditions.

Nevertheless, the findings should not be interpreted as proof that microbial profiles can reliably identify individuals in routine criminal investigations.

The experiments involved controlled transfer conditions, and classification performance varied depending on the sample type and processing method. Top-three accuracy also means the correct donor appeared among three ranked candidates, not necessarily as the leading prediction.

Further validation across larger populations, diverse environments and realistic evidence conditions would be necessary before broad operational application.

The study demonstrates the potential of microbial evidence while emphasizing the importance of evidence-specific validation and cautious interpretation of machine-learning results.