Ghost Lineages in the DNA

In a cold chamber beneath the Altai Mountains, a solitary tooth changed the conversation about our ancient cousins. The specimen, small and unassuming, comes from Denisova Cave, already the most consequential archaeological site for rewriting human evolution. But this molar did something rare. It yielded a genome, one with a clarity usually reserved for modern forensic labs, despite surviving roughly 200,000 years underground.

Relationships of Denisova 25 to archaic and modern humans. (A) Tree representing average population relationships inferred from nuclear genomes. Arrows indicate gene flow between populations, with darker arrows denoting those identified or examined in this study. Age estimates for high-coverage genomes are inferred from branch shortening (i.e., missing mutations in archaic genomes), with intervals corresponding to two standard deviations from the estimate as computed with a block jackknife. Population split time estimates are indicated by dotted lines. The split between the populations of Denisova 3 and Denisova 25 is estimated at 15,000 years (±2,000 years) before Denisova 25 lived, adjusted for the uncertainty in the age of Denisova 25. Neandertal-Denisovan and modern-archaic split times correspond to the range of estimates obtained with calibrations of the divergence in genomic regions that presumably evolve neutrally. The split between the population of the Denisova 5 Neandertal and other Neandertals was previously estimated³³. The tree is not to scale. (B) Bayesian phylogenetic tree of mitochondrial genomes of 7 Denisovans, 27 Neandertals, 64 modern humans and a hominin from Sima de los Huesos. The numbers at the nodes correspond to the posterior probabilities of the depicted branching orders. The Denisovan mitochondrial sequence of Denisova 19 is not included because it is identical to that of Denisova 21⁸. (C) Bayesian phylogenetic tree of the Y chromosomes of 3 Denisovans, 2 Neandertals, and 32 modern humans. The Y chromosomes of Denisova 4 and Denisova 8 were excluded from the analysis due to the lower coverage available, but positioned on the tree based on the sharing of derived alleles with the Y chromosome of Denisova 25. The length of the branches for Denisova 4 and Denisova 8 is inferred from the age of these individuals on the mitochondrial tree. The scale for the mitochondrial and Y chromosome phylogenies is indicated. Throughout the figure, Denisovans, Neandertals and modern humans are highlighted in light red, light blue, and light yellow, respectively.

This is only the second high-coverage Denisovan genome ever assembled.¹ The first, published more than a decade ago, belonged to an individual who lived tens of thousands of years later. Together, they sketch a picture of Denisovan history that is not tidy. Instead, it is layered, fractured, and threaded with encounters that shaped populations we still carry traces of today.