Around 41,000 years ago, a young woman died in a cold cave, high up in Siberia’s Altai Mountains. Scientists uncovered one of her pinky bones in 2008. From it, they extracted her DNA. And from that, they deduced that she belonged to a previously unknown group of ancient hominin, whom they called the Denisovans after the cave where the finger was found.

To this date, we have no idea what a Denisovan looked like. You can still hold every known Denisovan fossil—that pinky, a toe, and two teeth—in your hand. But we know so much else about them. We know almost every letter of their genome. We know that they diverged from their close relatives, the Neanderthals, around 400,000 years ago, and that both groups diverged from Homo sapiens around 600,000 years ago. We know that when our ancestors left Africa and spread into Asia, they encountered the Denisovans and had sex with them. We know that, as a result, Denisovan DNA lives on in people from Asia and Melanesia. One of these Denisovan genes provides modern Tibetans with a crucial adaptation that allows them to survive at high altitudes.

And now, thanks to work from Sharon Browning at the University of Washington, we know that Denisovan DNA entered the human gene pool on two occasions. Two separate groups of our horny, globe-trotting ancestors met with these mysterious hominins, and mated with them.

Browning developed a technique that compares the genomes of many modern people, and looks for stretches of DNA that are unusually varied, relative to their neighboring segments. These varied stretches are likely to have been inherited from ancient hominins like Neanderthals, Denisovans, or as-yet-undiscovered groups. Browning can then compare these segments to the genomes of those ancient hominins to work out exactly which group the DNA came from.

In the genomes of people from Japan and China, Browning found several segments that closely match the genome of the Denisovan woman whose pinky was found in the Altai Mountains. But to her surprise, she also found segments that were clearly Denisovan in origin, but were much weaker matches to the Altai female. “They’re close enough that we’re sure they’re Denisovan, but they’re not as close,” she says.

This suggests that modern humans inherited DNA from two separate groups of Denisovans. The way Browning sees it, the ancestors of today’s Melanesians encountered Denisovans as they expanded through South Asia and the Malay Peninsula. The two groups interbred, with the result that 5 percent of modern Melanesian DNA has a Denisovan origin.

At roughly the same time, the ancestors of today’s East Asians were taking a more northerly route through the continent, and met a different group of Denisovans, perhaps somewhere close to that Altai cave. Again, there was sex. Again, Denisovan DNA infiltrated the human genome.

But this contribution was much smaller, and today’s East Asians retain much less Denisovan ancestry than Melanesians do—just 0.2 percent of their genome. “Maybe it was a large group of East Asians meeting a small number of Denisovans,” Browning speculates. And complicating matters further, some Melanesians likely migrated north and mixed with East Asians, uniting the two waves of Denisovan ancestry in one lineage.

“Once I found two Denisovan waves, I thought we’d see something similar in the Neanderthals,” Browning says. To her surprise, she didn’t. People from Eurasia can trace around 2 percent of their DNA to Neanderthals, and it seems that this legacy comes from a single wave of interbreeding with one particular group—one that likely lived in the European part of their range.

That goes “somewhat against the grain of current knowledge,” says Emilia Huerta Sánchez, from the University of California at Merced. Some scientists have suggested that several groups of Neanderthals donated DNA to modern humans, which might explain why East Asians contain considerably more Neanderthal DNA than Europeans do. Browning and her colleagues argue otherwise, “and they have a considerable level of evidence to support their conclusions,” Huerta Sánchez says.

That seems counterintuitive, especially since we know that Neanderthals lived throughout Europe and Asia, whereas Denisovans have only been found in a single Siberian cave. But that picture might be deceptive. Scientists have sequenced the genomes of two Neanderthals—one from Vindija Cave, in Croatia, and another from the Altai region, in Siberia—and the differences between them are smaller than between the two Denisovan populations Browning identified in her work. “Maybe the Neanderthals were more nomadic and their populations were mixing a lot, while the Denisovans stayed in particular places and didn’t mix,” she says.

It’s also possible that Denisovans were more widespread than their limited bones suggest. “They may have overlapped in range with Asians and Papuans, as these populations had multiple encounters,” says Priya Moorjani, from University of California, Berkeley. “Since we only know about Denisovans from genetic material from one cave, any additional insights are exciting.”