Who else was here?

The genome from the finger bone described a population that split from the Neanderthal lineage hundreds of thousands of years ago and may have been widespread in Asia.

They contributed DNA to people living today: the paper that identified them put it at four to six per cent of Melanesian genomes, and later work found smaller amounts across east and south Asia and a high-altitude variant in Tibetans that came from them or from a population close to them. They were found in a genome before they were recognised in the ground.

The physical remains identified since are a handful of teeth and bone fragments, two jaws identified by their proteins, one from the Tibetan plateau and one from Taiwan, and a skull from Harbin.

The Tibetan jaw was the first to move the range. It was found at 3,280 metres on the Tibetan plateau, carries no usable DNA, and was assigned by the sequence of its collagen, which put Denisovans on the high plateau at least 160,000 years ago and a long way from the Siberian cave that named them.

The first genome came from a small amount of degraded DNA. In 2012 a method that separates the strands of the molecule before sequencing produced a genome from the same finger bone at a quality comparable to a modern one.

That allowed the population to be described rather than merely identified: small, with low genetic diversity, and separated from Neanderthals for hundreds of thousands of years.

It also sharpened the estimates of how much Denisovan ancestry survives in living people, and where.

Some analyses find segments in living West African genomes that appear to come from a population with no known fossils at all, though a 2023 model explains the same pattern with no archaic source. The study that produced the Altai Neanderthal genome also found a component in the Denisovan genome that may come from a further archaic population, or from older structure among their ancestors.

These are inferences from statistical patterns rather than from bones, which makes them weaker and still worth stating: the record of who was here is being read from living people as much as from the ground.

For fifteen years they were a genome with almost no body attached, and most of what was known about their range came from the DNA rather than from anything anyone had dug up. That shifted in 2025. Ninety-five proteins were recovered from a nearly complete cranium from Harbin, at least 146,000 years old, which had already been named as a species of its own; three of the amino acid variants in them are Denisovan, and the protein phylogeny puts the skull with the original Denisova genome. DNA from the tartar on its teeth, reported the same year, falls within Denisovan mitochondrial variation. It is the least fragmentary Denisovan anybody has.

The distribution of their DNA in living people suggests they were widespread in Asia, which makes the scarcity of fossils a statement about preservation and about where people have looked.

Sixty thousand years ago there were at least four: modern humans, Neanderthals, Denisovans, and the small-bodied population on Flores, whose bones are dated to between about 100,000 and 60,000 years ago. A fifth, named in 2019 from bones in a layer of a Luzon cave dated to 67,000 years ago, overlapped part of the same window.

Whether these are species, subspecies or populations is the species problem in a case where the evidence is fragmentary and the stakes are emotional. Modern humans, Neanderthals and Denisovans interbred where they met, which by one common definition makes them one species and by others does not.