Human prehistory is littered with tales of ancient trysts between the ancestors of modern-day people and populations of Neanderthals and Denisovans, our archaic cousins.
Evidence for this prehistoric intermingling comes largely from genomes sequenced from millennia-old Neanderthal and Denisovan bones from across Eurasia. As far as we know, neither cousin made it to Africa.
Computational geneticists Arun Durvasula and Sriram Sankararaman, from the University of California in Los Angeles, wondered, however, whether similar cross-species mixings did occur there.
The problem with trying to find out is that despite the rich fossil human record in Africa – once home to Homo sapiens, Homo naledi, Homo erectus and others – the continent has so far yielded no genomes from archaic humans.
To get around this, the pair came up with a computational method that relied solely on modern-day genomes, and the published genomes of Neanderthals and Denisovans.
Using publicly available data from west Africans – gathered as part of the 1000 Genomes Project – they measured how frequently a particular genetic sequence occurred in the modern population.
They repeated this at thousands of locations scattered throughout the genome and then plotted how often sequences occurred at low, intermediate and high frequencies.
Without any genetic in-flow from outside, there should be an even mix of frequencies across the spectrum. Some genes are present If the spectrum. Some genes are present in only a handful of people; others are shared by nearly everyone.
But the plot that emerged was a smile-shaped curve, with the high- and low-frequency ends turned upwards.
“This was a surprise,” says Sankararaman.
After ruling out errors and biases that might have crept into their analysis, they determined that the pattern they were seeing was the result of genes flowing in from an archaic relative.
“It turns out that this way of looking at the data, it looks very different if Africans did not have any archaic ancestry versus if they did,” he says.
Next, Durvasula and Sankararaman tested their data against a range of scenarios of how history played out to learn more about who the archaic relative was and when the intermixing occurred.
Neanderthals weren’t the source of the genetic in-flow, says Sankararaman. Instead, the analysis showed that a ‘ghost’ population was the likely source.
This population split from the ancestor of modern-day humans and Neanderthals between 360,000 years ago and one million years ago, before modern humans and Neanderthals split from each other.
The mixing between the ‘ghost’ population and humans took place sometime over the last 124,000 years. Mixing potentially occurring for an extended period of time, or on multiple occasions.
Remarkably, modern-day west Africans have retained between 2% and as much as 19% of the genome from this archaic population. “It’s quite a substantial contribution,” says Sankararaman.
The researchers identified a handful of genes whose ‘ghost’ hominin variant occur at unusually high frequencies in west African populations. Whether these variants confer particular benefits to the people who have them is something the pair will be investigating further.
“There would be some places in the genome where we do see these genes at high frequencies and many of those are likely to be adaptive,” says Sankararaman.
Whether the enigmatic ‘ghost’ hominin is one that’s been seen in the fossil record – Homo erectus, perhaps, or Homo naledi from southern Africa – remains a mystery. Given the lack of genetic information on archaic humans from Africa, it’s incredibly hard to pinpoint, says Sankararaman.
The finding could explain fossils that seem to have a mix of modern human and archaic features. Anatomically modern humans date back to around 200,000 years ago, but several fossils representing transitional forms or hybrids have been found dating to as late as 35,000 years ago in sub-Saharan Africa and the Middle East.
Whatever the specifics of how and when the populations intermingled, the story adds to the complex family history that has been seen in Eurasia with Neanderthals and Denisovans.
“This has been happening at different points in human evolution, at different geographic locations,” says Sankararaman. “Again and again, we have seen this.”
Recent research has suggested Africans also have some Neanderthal ancestry, challenging the assumption that only non-Africans bear signs of the Neanderthal lineage in their genomes. It indicates that ancient populations were migrating and mixing more than previously thought.
Sankararaman is keen to see what else might be hiding in the genomes of African populations. “Present-day African populations are extremely diverse,” he says. “The question is: do we see the same in southern Africa, in eastern Africa, in these highly diverse populations?”
The research is reported in a paper in the journal Science Advances.