The warship Vasa sank in Stockholm harbor in 1628, traveling just over 1,300 meters before it went down. When salvage crews raised it in 1961, they recovered skeletal remains along with it — clothing, artifacts, and bones that had spent 333 years submerged, then been excavated, reburied in soil, and in some cases stored in plastic. By the time forensic geneticists began working with those remains in earnest, the DNA was highly degraded, chemically damaged, and in places contaminated with mold.
Marie Allen, Professor of Forensic Genetics in the Department of Immunology, Genetics and Pathology at Uppsala University, has spent years developing methods to analyze exactly this kind of evidence — samples where standard approaches fail. Her research group focuses on improving sensitivity and discrimination in forensic DNA assays and on building strategies to counteract degradation, damage, mixtures, and contamination. The Vasa project applies those tools to one of the most complex historical identification challenges in Scandinavian forensic science.
Allen presented findings from the Vasa study at ISHI 36. The research was conducted in collaboration with colleagues at the Armed Forces DNA Identification Laboratory (AFDIL), the National Board of Forensic Medicine in Sweden, SNA International, and the Vasa Museum.
From Mitochondrial DNA to Whole Genomes
The investigation unfolded in stages. The first phase used mitochondrial DNA and traditional Sanger sequencing to re-associate fragmented remains and estimate how many individuals were represented. Early estimates from historical sources had placed the number at 25 to 30. DNA analysis brought that figure down. The current working estimate is a maximum of 15 to 16 individuals.
The more recent phase of the study focused on 15 individuals and applied next-generation sequencing methods, including low-coverage whole genome sequencing (WGS) and a targeted capture approach using the FORensic Capture Enrichment (FORCE) panel. The FORCE panel targets approximately 5,500 SNPs covering identity, ancestry, phenotype, and lineage markers — including X- and Y-chromosomal SNPs for kinship analysis. The panel was introduced in 2021 and was designed for exactly the kind of all-in-one forensic analysis that degraded historical samples require.
Even with severely degraded starting material, the FORCE panel enabled phenotype and ancestry predictions for nearly all 15 individuals. Allen described the collaborative approach to overcoming sample challenges: “we have been working on DNA analysis for many, many years and see that it’s fewer than [the original estimate]” — a result that required combining multiple methods across multiple institutions over an extended timeline.
A Finding That Made International Headlines
Among the results, one finding drew significant attention outside the forensic community. A skeleton that had been classified as male based on osteological analysis was genetically confirmed to be female. A second individual was also found to have a biological sex inconsistent with the prior osteological determination. Subsequent anthropological investigation supported the genomic findings in both cases.
“The DNA testing revealed that a male aboard this famed 17th-century warship was, in fact, a female, which has recently made headlines in international news.”
The result illustrates a practical use case that working forensic analysts will recognize: genomic sex determination as a check on or complement to skeletal analysis, particularly in cases where remains are fragmentary, poorly preserved, or subject to taphonomic changes that complicate morphological assessment.
Relatedness, Reconstruction, and What the Data Can Show
Beyond individual identification, the FORCE data were used to assess relatedness among the individuals on board. The analysis identified two individuals who may be siblings and two others — found in close physical proximity, wearing jackets made from the same cloth, and of similar age — who are believed to have been a couple. The clothing, like much of the ship’s contents, survived the centuries underwater and remains on display at the Vasa Museum.
The genetic data also fed into the museum’s physical reconstructions. Allen’s team provided predictions of eye color, hair color, and skin tone — information used to evaluate whether the facial reconstructions already on display at the museum were consistent with what the DNA showed. WGS data were used to analyze mitochondrial and Y-chromosome haplogroups and SNP-based ancestry. The 15 individuals showed European haplogroups; isotope analysis supported the interpretation that most likely grew up in various regions of Sweden or Finland. The majority were predicted to be blond and blue-eyed, with light to intermediate skin tones.
What Historical Casework Brings Back to the Field
Allen was direct about what draws her research group to conferences like ISHI: exposure to new technologies and new approaches to challenging samples. Her lab works across both historical DNA research and active forensic casework, including cold case investigations and missing persons work. The Vasa project is one data point in a broader research agenda focused on the kinds of evidence that conventional methods leave unresolved.
The abstract she presented at ISHI 36 closed with an observation that applies well beyond the Vasa: “Advancements in next-generation sequencing (NGS) have enabled the development of more sensitive techniques suitable for degraded DNA in historical remains… advanced tools such as MPS, WGS, and high-density SNP approaches now enable the resolution of a greater number of these long-standing cases, providing closure to victims and their families after years of uncertainty.”
The Vasa findings have been incorporated into an updated exhibition at Stockholm’s Vasa Museum, where the ship, its artifacts, skeletal remains, and reconstructions have been on public display since 1990. For Allen’s team, the work continues — whole genome sequencing data from the 15 individuals remains available for future genealogical analysis as methodologies continue to develop.