The third challenge is coverage — the number of times each position in the genome is sequenced. Higher coverage allows more confident base calls and more reliable genotyping. Many ancient DNA specimens, including several in the Chagyrskaya study, have low coverage — sometimes less than one-fold — which means that many genomic positions are read only once or not at all. The kinship analysis methods used by the team were adapted to work with low-coverage data, using probabilistic approaches that account for the uncertainty in individual genotype calls rather than treating each call as definitive.
The normalization of pairwise divergence (p0) used for kinship analysis is specifically designed for low-coverage ancient DNA scenarios. Rather than requiring confident genotype calls at every position, it works with the fraction of positions where two individuals show the same or different bases, treating each comparison probabilistically. This makes the method more robust to the specific challenges of ancient genomic data while still providing statistically reliable estimates of relatedness.
The sex determination method — comparing X chromosome coverage to autosomal coverage — is similarly adapted to low-coverage scenarios: rather than requiring full coverage of the sex chromosomes, it uses the ratio of sequence reads mapping to the X chromosome versus autosomes, which is informative even at low overall coverage.
The convergence of these multiple methods — sex determination, kinship analysis, mitochondrial analysis, heteroplasmy analysis — across a dataset of 13 individuals is what gives the father-daughter finding its evidential weight. No single method would have been sufficient. The conclusion emerges from the alignment of multiple independent lines of evidence, each with its own assumptions and potential error sources.
