Ancient DNA research has gone through a rapid expansion since the early 2010s. The number of ancient human genomes available for comparison has grown from dozens to tens of thousands. Sequencing technology has improved steadily. The statistical methods for ancestry decomposition, contamination estimation, and phenotypic inference have all been refined. The result is a field in which reanalysis of previously sequenced individuals — armed with better methods and better comparison data — regularly produces revised conclusions.
The Ötzi case is a prominent example of a broader phenomenon. The 2012 Keller genome was not poorly done science; it was state-of-the-art science for 2012. The 2023 Wang genome is state-of-the-art science for 2023. The difference between them reflects more than a decade of field-wide methodological development, not any failure of the earlier researchers.
What this trajectory also implies is that the 2023 findings are themselves provisional. They are the best analysis of the available material with the best current methods. Future improvements — in sequencing coverage, in phenotypic inference from ancient DNA, in the comparison dataset of ancient European genomes, in the ability to draw population-level conclusions from single individuals — will continue to refine the picture. The 90 percent Anatolian farmer figure comes with a ± 2.5 percent confidence interval, the phenotypic predictions are probabilistic, and the population-level interpretations are explicitly framed as requiring more individuals from the same region and period to test.
Johannes Krause noted in communications around the 2023 study that it remained unclear whether Ötzi was representative of people from his time and place, or whether he was an outlier in his unusually high Anatolian farmer ancestry. This is the right scientific posture: a single individual from a specific location and time period represents a data point, not a population.