Every person alive today carries DNA from extinct relatives whose fossils have never been found. That is the central claim of a study published July 30, 2026, in the journal Science, from researchers at the University of California, Berkeley, who developed a computational method capable of detecting ancient interbreeding between modern humans and unknown archaic populations, all without needing a single fragment of actual fossil DNA to work from.

Building a family tree without the bones

The research, led by graduate student Yulin Zhang and senior author Priya Moorjani, addresses a longstanding limitation in how scientists study ancient human interbreeding. Since the landmark sequencing of Neanderthal and Denisovan genomes years ago, researchers have understood that Homo sapiens interbred with these archaic relatives, leaving detectable genetic traces still present in people today. But that entire line of research has always depended on having actual ancient DNA recovered from fossils to compare against, and fossil DNA is exceptionally rare, requiring specific preservation conditions that most ancient remains never experienced.

To get around that limitation, the Berkeley team developed a new computational method called TRACE, short for TRacking Archaic Contributions via ARG Estimation. Rather than comparing modern genomes against ancient fossil samples, TRACE reconstructs genealogical relationships buried within hundreds of genomes from living people today, identifying DNA segments whose ancestry traces back further in time than the rest of the surrounding genetic material would suggest. As Moorjani described it, "these new computational methods that allow us to reconstruct genealogical relationships are really the next frontier in this field because they are allowing us to uncover hidden episodes from our past without requiring ancient DNA."

Confirming the method works before trusting its new findings

Before applying TRACE to search for entirely unknown ancestral populations, the researchers first tested it against genetic signals scientists already understood well. The method successfully identified DNA segments matching a sequenced Neanderthal genome, correctly recovering the well-established finding that Neanderthal ancestry accounts for roughly 1% of the modern human genome. It also correctly picked out known Denisovan DNA when researchers focused specifically on genomes from Asia and Oceania, regions where some present-day populations carry substantial Denisovan genetic ancestry. That successful validation against already-confirmed archaic contributions gave the research team confidence that whatever else TRACE detected in the broader genomic dataset likely reflected genuine, previously undetected ancient interbreeding events rather than a statistical artifact of the method itself.

A ghost lineage hiding across the entire human species

Once validated, TRACE revealed genetic regions that matched neither Neanderthal nor Denisovan reference DNA at all. The researchers concluded these unexplained segments came from two separate, previously unidentified archaic lineages that interbred with different human populations at different points in time. The first, which researchers term simply a "ghost" lineage, appears to have interbred with anatomically modern Homo sapiens in Africa more than 50,000 years ago, before modern humans undertook their major migration into Europe and Asia.

What makes this particular ghost lineage especially striking is how broadly its genetic signature has spread: the researchers found it present in both African and non-African populations alike, indicating the original interbreeding event occurred before humanity's final major dispersal out of Africa. The team estimates that DNA from this unknown population makes up roughly 0.5% to 1% of the genome of people alive today, a proportion comparable to the Neanderthal contribution already well established in the scientific literature. Based on the genetic divergence patterns detected, researchers believe this ghost lineage itself split from the ancestors of modern humans around 800,000 years ago, placing its origin roughly around the same period when the ancestors of Neanderthals and Denisovans began diverging from the broader human family tree.

A far older ancestor, discovered secondhand

The second newly identified population traces back considerably further in time and reached modern humans through an unusually indirect genetic pathway. Researchers call this lineage a "super-archaic" ancestor, tracing its origin to a hominin population that split off roughly 1.8 million years ago, making it dramatically older than either the Neanderthal or Denisovan lineages. Rather than interbreeding directly with modern humans, this super-archaic population appears to have interbred with Denisovans in Eurasia, likely more than 200,000 years ago. Denisovans then passed a portion of that super-archaic genetic material along to modern humans through their own later interbreeding with Homo sapiens, meaning this ancient lineage's DNA reached people alive today only after passing through an intermediary archaic population first.

Researchers detected this super-archaic signal specifically within genomes from populations in Oceania, and notably, only within the specific genomic regions already known to carry Denisovan ancestry, a pattern consistent with exactly the kind of indirect, two-step inheritance the researchers propose. Arjun Biddanda, a postdoctoral researcher at Johns Hopkins University and co-first author of the study, described why this particular finding felt especially significant to the research team: "The super-archaic finding is particularly exciting because it reveals genetic contributions from a human lineage that lived over a million years ago, despite the absence of any sequenced DNA from that population."

A plausible candidate identity, held loosely

While the researchers stopped short of definitively naming either newly identified lineage as belonging to a specific known hominin species, they offered a tentative candidate for the older, super-archaic population: Homo erectus, an early human species known to have lived across parts of Africa and Asia roughly between 110,000 and 1.89 million years ago. That timeline fits reasonably well with the roughly 1.8-million-year divergence date the genetic analysis produced, though researchers were careful to characterize this as a plausible hypothesis rather than a confirmed identification, since no actual Homo erectus DNA has ever been successfully sequenced to compare directly against the genetic signal TRACE detected.

Rewriting the shape of the human family tree

What both findings collectively reinforce is a picture of human evolutionary history considerably messier and more interconnected than the relatively clean, linear family tree many people still picture. Homo sapiens, Neanderthals, and Denisovans all descended from a common African ancestor, with modern human ancestors diverging from that shared lineage somewhere between roughly 550,000 and 765,000 years ago, followed by the Neanderthal and Denisovan lineages splitting apart from each other sometime later still. Layered on top of that already-branching structure, this new research adds at least two additional populations, one that interbred directly with early modern humans in Africa, and one considerably older population whose genetic legacy reached us only indirectly, through Denisovans, adding yet more knots to an evolutionary tree already known to be far more tangled than scientists assumed even a decade ago.

A method with reach well beyond human genetics

Beyond the specific findings involving human ancestry, Moorjani pointed to TRACE's broader potential utility across evolutionary biology generally, noting the method "should also work with other species, allowing us to also uncover really different patterns across the tree of life." Because TRACE doesn't require the extraordinarily rare preservation conditions needed to recover usable ancient DNA directly, the same underlying computational approach could plausibly be applied to study hidden interbreeding history across a considerably wider range of species than fossil-DNA-dependent methods have ever been able to investigate.

For human evolutionary research specifically, the discovery of these two ghost lineages suggests that even after decades of intensive study following the original Neanderthal and Denisovan genome breakthroughs, the full cast of ancient relatives who shaped modern humanity's genetic makeup likely remains incompletely known, waiting to be uncovered not necessarily through the discovery of new fossils, but through increasingly sophisticated ways of reading the genetic history already recorded, quietly, inside the DNA of people alive today.

---

*Sources cited in this article include the peer-reviewed study "Recovering signatures of archaic hominin introgression using ancestral recombination graphs," published July 30, 2026, in Science, and reporting from ScienceDaily, Phys.org, Smithsonian Magazine, SciTechDaily, Sci.News, Live Science, and Daily Galaxy covering research led by Yulin Zhang and Priya Moorjani at the University of California, Berkeley. All figures reflect reporting available as of August 8, 2026.*