Logo: to the web site of Uppsala University

Mitochondrial DNA (mtDNA) analysis plays a specialized role in forensic applications, overcoming certain limitations of autosomal DNA markers. The high copy number and uniparental inheritance pattern of mtDNA are advantageous in cases involving shed hairs and aged skeletal elements, especially decades-old missing persons cases. Though the discriminatory power of mtDNA is limited by common haplotypes, next generation sequencing (NGS) offers feasible access to entire mitochondrial genome (mitogenome) data that can provide increased resolution of common haplotypes to unique sequences. The primary implementation challenge of mitogenome analysis is a lack of forensic-quality reference data, which are required to determine the evidentiary weight of a match. A better understanding of NGS methods and data analysis is also necessary to ensure the generation of reliable mitogenome data. Furthermore, appropriate quality control (QC) measures must be established as analysis can be complicated by nuclear mtDNA segments (NUMTs), misalignment of homopolymer regions, sequencing errors, and other artefacts. Including such false variants in mtDNA haplotypes can lead to erroneous conclusions based on misinterpreted data.

This thesis aimed to address the implementation challenges of mitogenome analysis and facilitate the transition to NGS in forensic laboratories. Paper I assessed the feasibility of generating forensic-quality mitogenome data from whole genome sequencing (WGS) data, which are valuable sources of mitogenome haplotypes for population studies.  Due to NUMT interference, a 10% variant frequency threshold was necessary to produce haplotypes consistent with high-quality mitogenome datasets. Since length heteroplasmy (LHP) can also complicate mtDNA data analysis, Paper II characterized LHP in data generated on two NGS platforms as well as with Sanger-type sequencing. Different patterns of LHP were observed across sequencing technologies, further supporting current guidelines to ignore LHP in database searches and match comparisons in forensic analyses. Phylogenetic information can provide a valuable QC check of mtDNA data, identifying errors like artificial recombination and phantom mutations. Therefore, three haplogrouping tools were examined in Paper III, comparing their ability to predict an accurate haplogroup based on different mitogenome target ranges. The tools performed similarly, but EMPOP’s SAM2 algorithm produced more precise haplogroup predictions than the other two tools regardless of phylogeny or interpretation range. Building upon the previous three studies, Paper IV characterized 934 forensic-quality Swedish mitogenomes from a population genetics perspective. The complete mitogenome data demonstrated high haplotype diversity (0.9996) with a random match probability of 0.15%. In summary, these papers combine important insights to facilitate the application of mitogenome NGS analysis in forensic laboratories.