The analysis of individual skin flakes represents a powerful method for detecting DNA from a person of interest (POI) in highly complex mixed traces or in traces where the DNA of a POI is present only as a minor component. This investigation of skin flakes is frequently applied in the context of cold cases. However, due to the inherently low quantity of DNA typically present in such samples, possibly amplified by storage-related DNA degradation, the resulting DNA profiles are often incomplete. Furthermore, these profiles commonly exhibit strong stochastic effects, including peak imbalances, allele or locus drop-outs, and drop-ins and - due to the frequently used low template PCR protocols - increased stutter peaks must also be expected. The biostatistical analysis of such partial profiles using binary statistical models is only possible to a limited extent, as these approaches cannot incorporate drop-ins and generally require fully typed STR-loci for calculation. A potential alternative is the use of fully continuous probabilistic genotyping software such as EuroForMix (EFM), which includes stochastic effects during the modeling process and therefore allows the full evidential information to be considered. However, it remains unclear to what extent EFM can sufficiently model strong stochastic effects. In this study, measurement data from skin flakes of known individuals originating from cold-case investigations were collected and evaluated using EFM. Samples of varying quality were intentionally selected based on criteria such as DNA degradation and DNA quantity in order to represent a wide spectrum of stochastic effects. The results of this study show that EFM can generally provide more informative likelihood values than binary models. However, when stochastic effects become too strong, reliable modelling is no longer possible, leading to likelihood values that cannot be considered robust for forensic interpretation.

Authors

• Jan Fleckhaus (Institute of Legal Medicine, Munich, Germany)
• Katja Anslinger (Institute of Legal Medicine, Munich, Germany)