Many forensic genetic trace samples (e.g., telogen hair) are of too low quality to obtain short tandem repeat (STR) DNA profiles. Instead, performing shotgun DNA sequencing of such samples can provide valuable information on single nucleotide polymorphism (SNP) for human identification. As a result, shotgun sequencing is starting to gain more attention in forensic genetics, but shotgun sequencing is not perfect. 

In most cases, the reference samples are of high quality with measurably low genotyping error probability. This is in contrast to the trace samples that can be of low qualities in many different ways and the sample-specific genotyping error probability can be difficult to determine. 

Statistical models are needed to correctly interpret such evidence, including properly accounting for different (asymmetric) genotyping error probabilities (such as those found when comparing a low-quality trace sample to a high-quality reference sample) and handling unknown genotyping error probabilities. 

Dealing with asymmetric and unknown genotyping error probabilities are presented based on the wgsLR model that is one such model to interpret shotgun sequencing forensic genetic evidence. 

Authors

• Mikkel Meyer Andersen (Department of Mathematical Sciences, Aalborg University, Denmark and Section of Forensic Genetics, Department of Forensic Medicine, University of Copenhagen, Denmark, Denmark)