The Ghost In Our Genes: Scientists Discover Mystery Ancestor Hiding In Human DNA

Most Humans Carry DNA From an Unknown Ancient Relative, Study Finds

A genetic ghost appears to be hiding across the human genome. Not a Neanderthal, not a Denisovan, but something else, an unknown ancient relative whose traces turned up in every population sampled in a sweeping new analysis, from sub-Saharan Africa to Southeast Asia to the Pacific islands. The mixing likely occurred before any group migrated out of Africa, according to the study’s model.

Scientists have long known modern humans carry small amounts of DNA from Neanderthals and Denisovans, ancient relatives whose bones and genomes have been studied for years. This new study, published in the journal Science, points to a third, unnamed group, a “ghost” lineage for which no skeleton or genome has ever been found.

To find these hidden traces, researchers built a computational tool called TRACE, short for TRacking Archaic Contributions via ARG Estimation. Rather than relying on ancient bones or a reference genome from the mystery species, neither of which exist, TRACE reads the family tree inside the DNA of people alive today, hunting for genetic code that looks too old and too strange to belong to our direct lineage.

A New Tool Spots Ancient DNA With No Bones Required

Studying ancient interbreeding typically requires a sequenced genome from the species itself, or a group of modern humans with zero ancestry from it, serving as a control. For Neanderthals and Denisovans, scientists had both. For this ghost population, neither exists.

TRACE sidesteps that problem, mapping every time DNA was shuffled and recombined across thousands of generations. DNA inherited from a deeply ancient outside group leaves a distinct signature: it sits on unusually long branches of the family tree, a pattern researchers interpret as evidence of interbreeding with a separate lineage.

In simulations using genealogical trees where the true answer was known, TRACE achieved 92% accuracy and a false discovery rate below 0.25%, though it did not recover every genuine archaic segment. Applied to genome sequences from the 1000 Genomes Project, a large international database spanning dozens of populations, the tool confirmed established findings: Europeans, East Asians, and South Asians each carry roughly 0.8 to 1% Neanderthal ancestry, with higher Denisovan ancestry in East and South Asians, matching prior research and lending credibility to TRACE’s outputs.

The Mystery DNA Makes Up Roughly 1% of the Genome

Beyond the familiar Neanderthal and Denisovan signals, TRACE flagged something unexpected: a genetic source matching no known ancient group, making up between 0.5 and 1.1% of the genome. West Africans, East Africans, Europeans, and Oceanians all carry it, and most ghost segments in non-African populations are shared with sub-Saharan Africans.

These segments showed nearly equal genetic similarity to both Neanderthals and Denisovans, rather than clustering with either, a signature of a third lineage related to both but distinct from them. This ghost population is estimated to have split from the Neanderthal-Denisovan lineage more than 500,000 years ago, though the figure is model-based rather than directly observed. Its identity remains unknown; the authors suggest it could plausibly correspond to Middle Pleistocene human groups or African Homo heidelbergensis populations, though no definitive identification is possible from genetic data alone. Ghost segments also tend to be shorter than Neanderthal or Denisovan ones, consistent with an older mixing event having more time to fragment. Simulations lacking a ghost lineage produced almost no signal, while simulations including one matched the real-world data closely.

So-Called “Empty” DNA Deserts Turn Out to Hold Ghost Ancestry

One striking finding involves regions labeled “archaic deserts,” long stretches where Neanderthal and Denisovan ancestry is essentially absent, often interpreted as selection purging harmful ancient DNA.

TRACE complicated that story. All five shared deserts examined contain substantial ghost ancestry. A desert on chromosome 7, containing a gene linked to language and speech, shows a ghost peak of 13.3%; a desert on chromosome 3 has a ghost peak near 20%. If these regions rejected all ancient DNA on principle, ghost ancestry would not survive there either. Instead, whatever selection occurred appears specific to Neanderthal and Denisovan DNA, not a blanket rejection of outside ancestry.

For people from Oceania, including Papua New Guinea, Vanuatu, and the Santa Cruz Islands, the story goes deeper. Analysis of 92 Oceanian genomes uncovered signs of a “super-archaic” lineage, estimated to have diverged from modern humans’ ancestors around 1.77 million years ago. The signal appeared within DNA already identified as Denisovan, consistent with Denisovans having interbred with an even older group and passed that ancestry along, though the paper treats this as a hypothesis, not a proven pathway. One candidate, the paper notes, is Homo erectus, a possibility earlier studies have also raised, though it remains speculative. Genes here relate to immune function and vitamin D metabolism, but any measurable biological effect has not been established.

Human history looks far messier than a clean family tree suggests. Modern humans did not simply march out of Africa and replace everyone else; the genetic record points to repeated mingling with relatives spanning millions of years. TRACE offers a way to read that history without a fossil or genome from every participant, and as the methods improve, the picture should only get sharper.

---

---