Neanderthal in CAve
© Yulliii/Shutterstock
When the Neanderthal genome was first sequenced in 2010 and compared with ours, scientists noticed that genes from Homo neanderthalensis also showed up in our own DNA. The conclusion was inescapable: Our ancestors mated and reproduced with another lineage of now-extinct humans who live on today in our genes.

When the Denisovan genome was sequenced soon after, in 2012, it revealed similar instances of interbreeding. We now know that small populations from all three Homo lineages mixed and mingled at various times. The result is that our DNA today is speckled with contributions from ancient hominin groups who lived alongside us, but did not survive to the present day. Genes from Denisovans and Neanderthals are not present in everyone's DNA - for example, some Africans have neither, while Europeans have just Neanderthal genes. But, these genetic echoes are loud enough to stand out clearly to scientists.

On one level, it's not shocking that DNA from other human groups resides within us. H. sapiens today is the result of millions of years of evolution; we can count numerous species of ancient hominin among our ancestors. But the Neanderthal and Denisovan contributions to our genetic makeup happened far more recently, after H. sapiens had already split from other human groups. Those interbreeding events, also called introgressions, did not create a new species of human - they enriched an already existing one. Some of the traits we acquired are still relevant to our lives today.

"There's a lot of evidence for some type of introgression from ancient hominins into modern humans, particularly modern humans out of Africa," says Adam Siepel, a computational biologists at the Cold Spring Harbor Laboratory. "I don't think there's any real question among experts in the field as to whether the evidence overwhelmingly supports that event."
Finger Bone Fragment
© Thilo Parg/Wikimedia CommonsReplica of a Denisovan finger bone fragment, originally found in Denisova Cave in 2008, at the Museum of Natural Sciences in Brussels, Belgium.
Some evidence also suggests that there may be more than two additional human groups lurking in our DNA, what researchers sometimes call "ghost lineages." Modern humans living in Africa may have interbred with one or more hominin species there, resulting in even more addition to our current DNA. And a recent study of modern-day Indonesians suggests that what we call Denisovans was actually three separate groups of hominins, at least one of which can be thought of as its own species. The ancestors of Asians and Melanesians mated with at least one of these groups, and possibly more.

Hunting for Ghosts

Scientists have begun to unravel our tangled past in a few different ways. For groups like the Neanderthals and Denisovans, where their DNA still survives in some fossils, it's been easiest. After sequencing their genomes, archaeologists simply compared them with our own. Finding stretches of our DNA that look strikingly similar to those from an ancient group is a strong indication that our ancestors interbred with them at some point.

Those genes could have come from even further back, of course, from the last common ancestor we shared with Neanderthals and Denisovans. But the split happened so long ago that most regions of the DNA we share with other human groups has seen mutations that make them look distinct in each group. Any genes that seem to be a direct match, then, are a strong indication of interbreeding.

For the majority of ancient hominin species that we know of, though, we have bones but no DNA. The molecules that make up the delicate double helix break apart over time, especially in the hot and humid environments where most of our ancestors lived. But archaeologists have unearthed faint echoes of additional ancient hominins in our own DNA by running genomic data through statistical algorithms meant to pick out telltale variations. Several papers have suggested that hidden ancestors lurk in our genomes. But the margins for error are large.

Neanderthal skull
© Le Club de MediapartA Neanderthal skull discovered in 1908 at La Chapelle-aux-Saints (France).
The mathematical models scientists use to find ghost lineages in genetic data are complex, but usually boil down to a search for clusters of genes in specific groups of humans today that set them apart from other modern human populations. A team of researchers did just this recently with a collection of 161 genomes from modern-day Indonesians. They compared this genetic information with the genomes of Denisovans and Neanderthals, as well as other humans today.

The researchers found signatures of Denisovan DNA in the Indonesians - no surprise - but also saw hints that the ancient hominins were actually three distinct groups. They shared a common ancestor but began to drift apart genetically as they spread across Asia and the Pacific. One of the groups was as different from the Denisovans found in Siberia as it was from Neanderthals. The researchers say that means it likely deserves to be called its own species, though they don't have the fossils to make the designation official.

"If we're going to call Neanderthals and Denisovans by a unique name, which we do, then we should probably call this other group by another name," says Murray Cox, a computational biologist at Massey University in New Zealand, a co-author of the study.

New finds may give us tangible evidence of this Denisovan cousin, but for now its only remains are snippets of DNA tucked inside the genomes of some modern human populations.

Those Denisovan-like ghosts are not alone, though. Other researchers looking at the DNA of African hunter-gatherers today have used similar methods to find what they say is evidence that the ancestors of those groups mated with other hominins on the continent tens of thousands of years ago.

Perhaps the largest study, published in 2011, looked at the DNA of 61 Africans from the Mandenka, Biaka and San tribes. They compared these genomes to two models of human populations, one of which assumed the Africans' ancestors interbred with ancient human groups, and one which didn't. The model that included gene flow from archaic hominins produced results that more closely matched up to actual human populations in the region.

Based on their modeling, the researchers say that around two percent of the African genomes they sequenced came from a mysterious group of ancient hominins. The two interbred somewhere around 35,000 years ago.

Another study looking at a gene, called MUCL7, that encodes for a protein in our saliva. They found further evidence of a ghost lineage in Africans. The MUCL7 gene has a few variants in humans today, and the researchers say that one can be traced back to an ancestor that was something other than H. sapiens. A paper examining the genomes of 15 people from the Hadza, Sandawe, Baka, Bakola and Bedzan tribes also found evidence of interbreeding, though again they were unable to say whom it might have been.

Understanding Our Inheritance

The genes that many of us have inherited from other human groups remain, in many cases, active in our bodies today. Genes from Neanderthals and Denisovans affect the functioning of our immune systems today. Neanderthal genes have been suggested to affect both the body's keratin (which makes up our hair, among other things) and how we react to UV light. These genes were likely picked up after humans migrated north from Africa to Europe, where trysts with our Neanderthal cousins may have helped fast-track our own adaptations to the cold climate.

A unique gene from Denisovans has also been found in modern-day Tibetans, high in the Himalayas. Known as EPAS1, it alters how their bodies produce hemoglobin, a protein in red blood cells that carries oxygen. The gene helps move oxygen around their bodies more efficiently at altitude, and has likely been instrumental in allowing them to colonize the miles-high world of the Tibetan Plateau.

Other genetic endowments have proven more problematic. Some genes from Neanderthals have been implicated in depression and myocardial infarctions, as well as allergies. Where some genes from ancient species help our immune systems, others may predispose us to disease: lupus, Crohn's disease and type-2 diabetes, among others.

Like many family legacies, our extra-sapien genetic inheritance is complex. Scientists have more work to do before they understand the true implications of our multi-species past, including the ways that it continues to shape our present. Indeed, even finding evidence for ghostly genes can be problematic.

Telling Ghosts from Phantoms

The evidence for ghost lineages is building, but when scientists must rely solely on genetic evidence, it is difficult to make an airtight case for the existance of another species. There are numerous factors that can muddle the stories our genes tell, making it difficult to tell a true signal of an ancient population from background genetic noise. What kind of evolutionary pressures a population is under, the size of that population, to what extent it has mixed with other human populations and things like genetic bottlenecks - an event where just a fraction of a population survives to produce offspring - all add uncertainty to geneticists' work. What appear to be ghosts are sometimes no more than metaphorical curtains blowing in the wind.

A recent paper in Nature Communications illuminates some of the shortfalls of this method. In it, three researchers from Spain and Estonia say that they've found evidence of a ghost lineage in the DNA of people of Asian and Oceanian descent based on a complex statistical analysis and machine learning algorithms.

They started by constructing an artificial genome using a computer program that simulates genetic information and then paired it with demographic models of ancient humans. This gave them an "in silico" genome that they could compare with real ones, says study co-author Oscar Lao, a population geneticist at the Centre for Genomic Regulation in Barcelona. They ran these simulations millions of times.
Denisova Cave in Russia
© Wikimedia CommonsThe opening to Denisova Cave in Russia’s Altai mountains. The first Denisovan remains to be found came from the cave, in addition to bones from Neanderthals and modern humans.
Where the two matched up, they then looked at what in the artificial genome's evolutionary history led to that cluster of genes, with the implication that what created that DNA in the computer genome may also have happened in real life. Averaging the results over millions of simulations gave them what they say is an approximation of a real genome.

They then attempted to fit the data into a few different scenarios describing the interactions between various lineages of ancient humans. The one that fits best, they say, is a model that includes a heretofore unknown group of ancient hominins interbreeding with modern humans. These people would have been a hybrid of Neanderthals and Denisovans, the authors say.

It would also explain why Asians today have more Neanderthal DNA than Europeans do - not only did their ancestors mate with the Neanderthals at some point, they also interbred with this hybrid group in Asia. The work is another reminder that hominins were a diverse bunch, says Lao.

"What it's suggesting is that the picture of archaic genetic diversity was quite complex," he says. "It was not only Neanderthals waiting until anatomically modern humans appeared, but there were other archaic populations from which we don't have archaic remains so far."

It's a compelling story, boosted by the discovery last year of a first-generation Neanderthal-Denisovan hybrid in Denisova cave in Siberia. But other researchers aren't convinced.

Ancient Bones
© Max Planck InstituteResearchers uncovered the first direct evidence of hominin interbreeding in the bone fragments of a 13-year-old girl from Russia’s Altai Mountains.
There are just too many unknowns at the moment for us to run simulations of this kind, Siepel says.

"It's a question where you have to write down a fairly complicated model to distinguish between those scenarios and that model is almost certainly unrealistic in many respects," he says.

Sharon Browning, a research professor of biostatistics at the University of Washington, agrees.

"The reality of human history is pretty complex," she says. "If you simplify too much and don't capture the right aspects of what really happened then you're going to be comparing different models, all of which are wrong."

Browning also studies the genetics of ancient hominins, and she says her own work found no evidence of the population that Lao and his colleagues saw.

Instead, a recent paper from her lab looking at traces of Denisovan DNA in modern genomes found two distinct Denisovan-like groups. One looked genetically similar to the single Denisovan genome we've sequenced, from Siberia. The other group looked to be only distantly related, though.

It's a finding that matches up, at least in broad outline, with the recent suggestion that the Denisovans were more than one species. If so, both Browning and Cox may have found evidence of the same ghost lineage distantly related to Denisovans. Or perhaps they each picked out different groups entirely, hinting that there was much greater diversity among ancient hominins than we realize.

More to Come?

As scientists claim evidence of unknown populations in our shared past, it's reasonable to wonder how many more there might be to find. What evidence we have so far of ghost lineages amounts mostly to hints, without (by definition) solid proof to back them up. Indeed, we may have already hit the limits of what is observable when it comes to picking out ancestors in our DNA. The Neanderthal and Denisovan additions amount to just a fraction of the genomes of some modern human populations.

"Any other introgression that's out there is going to be quite a small proportion [of our DNA]," Browning says. Most of the DNA segments that look like they came from another species have probably been found, she says. "There's not a lot left to work with that might be something else."

But future archaeological finds - providing real, hard evidence - could surprise us, especially if we can extract DNA from the remains. The evidence for Denisovan DNA in our genome, for example, would likely be quite tenuous had we not been able to sequence their DNA, Siepel says.

"I'm sure that would be very controversial, if we were just trying to hypothesize the existence of that species from the introgressed fragments in modern humans," he says.

Future finds may turn today's ghost lineages into tangible species, once-living and breathing humans who merged their genetic histories with ours. Their bodies may be gone, but their DNA lives on.