With highly degraded ancient DNA, the challenge is that the fragments you can recover are short. A short fragment that matches between two individuals might be shared ancestry — or it might be random coincidence, particularly if both individuals belong to a relatively small population where many people share common ancestry from the distant past.
This is where the comparison group becomes critical. Willerslev’s team compared the Sitting Bull hair sample not just with Ernie LaPointe but with a group of other Lakota Sioux individuals. This allows the researchers to ask: given the patterns of genetic sharing that exist across this population, does LaPointe’s sharing with the hair sample look like what you would expect from random common ancestry within the population — or does it look like something closer and more specific?
The probabilistic method developed for this study addresses exactly that question. It does not require pristine, high-coverage DNA sequences. It works with the degraded, short-fragment data that is available, and it attempts to estimate the probability that the observed patterns of sharing are consistent with the claimed relationship rather than with no specific relationship.
The fourteen-year timeline from the initial access to the hair around 2007 until the publication in 2021 reflects how difficult this technical problem was to solve. Existing ancient DNA methods at the time of the initial access were not capable of producing reliable kinship assessments at three generations of separation from highly degraded material. The team had to develop new approaches.
The published study’s methods section, which was not fully accessible during the research for this article due to database access restrictions, would contain the specific statistical details: the exact method for comparing short degraded fragments, the comparison group composition, the threshold criteria for concluding that the relationship is supported, and the statistical confidence of the result. What the publicly accessible reporting on the study consistently conveys is that the results “confirmed” or “supported” the claimed great-grandson relationship — that the patterns of genetic sharing between the hair sample and LaPointe’s DNA were not consistent with the degree of relatedness you would expect from random common ancestry within the Lakota Sioux population.