Berkeley Scientists Found Two Human Ancestors Nobody Has Ever Seen — Hiding in Our Own DNA

Scientists at the University of California, Berkeley have identified two previously unknown human ancestor populations — groups that left no bones, no tools and no fossils anyone has found, but whose DNA is still circulating in living people today.

The work, published in Science on July 30, 2026, was led by Priya Moorjani, an associate professor of molecular and cell biology at Berkeley, with graduate student Yulin Zhang and postdoctoral researcher Arjun Biddanda of Johns Hopkins University. The team built a computational method they call TRACE — TRacking Archaic Contributions via ARG Estimation — that reconstructs genealogies across the genome rather than comparing modern DNA against a reference sequence from a known archaic species.

That distinction is the whole point. The standard way to detect ancient interbreeding requires a sequenced genome from the archaic group in question. Neanderthal and Denisovan DNA were found because researchers had physical remains to sequence. TRACE instead looks for stretches of the genome whose inheritance histories are too deep and too divergent to belong to any known lineage, which lets it flag populations no one has ever dug up.

The first of the two is a "ghost" lineage that separated from the ancestors of modern humans roughly 800,000 years ago and then, more than 50,000 years ago, interbred with humans in Africa. It contributed about 1 percent of the modern human genome — a contribution comparable in size to the Neanderthal share carried by people of non-African descent.

The second is older and stranger. TRACE detected a "super-archaic" population dating back approximately 1.8 million years, which did not mix with our ancestors directly. It interbred with Denisovans more than 200,000 years ago, and its DNA reached modern humans secondhand, through later Denisovan admixture. The surviving fraction is small, but its presence means the Denisovans were themselves already a genetic mosaic before they ever encountered us.

"Genealogies preserve a record of our evolutionary past. TRACE reconstructs those histories across the genome," Moorjani said.

The cumulative effect of these findings is to keep pushing human ancestry away from the tidy branching tree of textbook diagrams and toward something the authors describe as more like a complex web — a set of populations that repeatedly separated, drifted apart for hundreds of thousands of years, met again and exchanged genes. Every human alive is carrying fragments of groups that were distinct enough to be considered separate populations and that vanished without leaving a skeleton behind. The record of their existence is not in the ground. It is in us.