On a windward beach in the Solomon Islands, a young woman rolls a small saliva tube between her fingers while a research assistant labels it with a study code. The tube will travel thousands of miles, eventually landing in a freezer at Yale University, where its contents will join a dataset that is rewriting how researchers think about human evolution. Like most people in the islands, the woman carries a small but functional inheritance from an extinct branch of humanity: DNA from a group called the Denisovans, who vanished from the Earth tens of thousands of years ago.
That inheritance is no longer a curiosity. A study published on June 11, 2026, in the journal Science shows that Denisovan DNA is not just sitting quietly in Oceanian genomes. It is actively regulating the immune system, switching genes on and off in ways that may still shape who gets sick, and from what [1][3]. The full paper, by Patrick F. Reilly and colleagues, is available under the DOI 10.1126/science.adr6749 [3]. Its conclusions are a quiet earthquake for a field that, for a decade, has been cataloguing which stretches of the modern human genome came from which archaic ancestor. The new work goes further. It shows what those stretches are actually doing.
Who were the Denisovans?
The Denisovans are a sister group to the Neanderthals, an extinct lineage of archaic humans who lived across Asia from roughly 200,000 to 32,000 years ago [6]. Researchers first identified them in 2010, from a fragment of mitochondrial DNA pulled out of a tiny finger bone excavated in 2008 from Denisova Cave in the Altai Mountains of Siberia. The bone belonged to a juvenile who had lived more than 50,000 years ago.
Since then, fossil and genetic evidence has placed Denisovans in an unexpectedly wide geography. Their DNA has been recovered from Baishiya Karst Cave on the Tibetan Plateau, from Tam Ngu Hao 2 Cave in Laos, from a jaw in the Penghu channel near Taiwan, and most recently, in June 2025, from a Harbin cranium in Manchuria [6]. They were not a small, peripheral population. They were a wide-ranging Asian hominin, and modern humans met them, repeatedly, on the way out of Africa.
Most non-African humans alive today carry a small percentage of Neanderthal DNA. Denisovan ancestry, by contrast, is patchy. Mainland Asians and Native Americans carry roughly 0.2%. Melanesians, Aboriginal Australians and the Aeta and Mamanwa groups of the Philippines carry closer to 5%, the highest Denisovan contribution found in any modern population [6]. Other estimates put the average Denisovan share in Oceanian genomes a little lower, near 2%, depending on the reference panel and the method used [6]. The exact percentage matters less than the underlying signal: the Pacific holds a Denisovan inheritance that almost nowhere else does, and almost nobody had looked at it carefully until now.
A new map of Oceanian ancestry
The new study, led by Serena Tucci, an assistant professor of anthropology at Yale and principal investigator of the Yale Human Evolutionary Genomics Laboratory, sequenced the genomes of 177 individuals from 12 distinct populations across Near Oceania, the region encompassing Papua New Guinea, the Bismarck Archipelago and the Solomon Islands [1][3]. To put those sequences in context, the team combined them with 1,284 previously published genomes from around the world [1][4]. Near Oceanic populations are the descendants of one of the earliest human dispersals out of the Neanderthal-Denisovan homelands, with ancestors who reached the region at least 45,000 years ago [1][5].
The headline finding is striking. Previous work had documented at least one major pulse of Denisovan interbreeding into Oceanian ancestors. The new analysis pushes that count up to at least three distinct Denisovan groups [1][3]. The Pacific, in other words, was not a single encounter. It was a series of them, scattered across deep time, with different Denisovan populations each leaving their own genetic fingerprint.
"The drastic underrepresentation of Oceanians limits our understanding of human evolution and could exacerbate health inequalities as genomic research is used to develop novel medical treatments," Tucci said [1]. The point is also practical. Most large-scale human genetics studies have leaned heavily on European-ancestry populations, and that bias is now baked into the reference panels used by ancestry tests and drug-response studies [4]. A clearer picture of Oceanian genomes is, among other things, a more honest picture of human biology.
The team also identified a strong population bottleneck in the ancestors of present-day Near Oceanians, a contraction in genetic diversity that is visible in the shape of every modern genome from the region [3]. Combined with three separate Denisovan introgression events, the result is a population history of unusual structure, made up of long isolation punctuated by brief, consequential meetings.
From correlation to function
For most of the past decade, the story of archaic DNA in modern humans has been a story of correlation. Researchers could look at a genome and say: this stretch looks Denisovan. They could, in some cases, say that the same stretch had been favoured by natural selection. But showing that the stretch actually did something, that it changed a behaviour of a real, living cell, has been much harder.
The Yale team set out to do exactly that. Using a technique called a massively parallel reporter assay (MPRA), they synthesised thousands of archaic DNA sequences and tested, in parallel, whether each one altered the activity of nearby genes [1][3]. The MPRA approach lets researchers ask, of every Denisovan variant, "is this one a switch, and which way does it flip?" The answer, the team reported, was that more than 3,100 of those archaic variants measurably tune gene expression up or down [1][3][4][5].
"With this study we have moved beyond simply 'resurrecting' this DNA to showing how it actively turns genes on and off, which is game-changing," Tucci said. "This DNA is not just a remnant of ancient liaisons; it continues to influence our biology today" [1][2].
That conclusion reframes the meaning of carrying archaic DNA. A variant is, in many cases, a working part of the cell, a living piece of machinery rather than a passive relic. The funding for the work came from the National Institute of General Medical Sciences and the National Human Genome Research Institute, both part of the U.S. National Institutes of Health [1].
The immune system, with Denisovan software
The bulk of those 3,100-plus working variants cluster in a familiar place: the interferon-gamma signaling pathway, a cytokine cascade that is one of the body's core defences against viruses and bacteria [1][2][5]. The same variants that flip gene expression also tend to sit in immune-related regulatory regions, and they tend to be the variants that have been most strongly favoured by selection in modern Oceanian populations.
For Patrick Reilly, the study's first author and an associate research scientist in Tucci's lab, this is the part of the result that most clearly links the deep past to the present [1]. "DNA from extinct hominins, Denisovans and Neanderthals, helped facilitate human adaptation to diverse environments that people encountered as they migrated into this region of the world," he said. "Pathogens are one of the strongest selective pressures, environmental factors that affect our ability to survive, throughout human evolution. We find evidence that genes inherited from Denisovans bolstered immunity to viruses and bacteria ancient humans encountered in Near Oceania" [1][3][5].
The trade-off shows up in modern disease patterns. Steven Reilly, an assistant professor of genetics at Yale School of Medicine and a co-author of the study, noted that the immune tuning is not free [1]. "We found thousands of archaic variants that tune genes up or down, concentrated in immune and antiviral pathways," he said. "Neanderthals and Denisovans had adapted to life outside Africa over hundreds of thousands of years, and we inherited some of those genetic programs and co-opted them. Tens of thousands of years later, this DNA may still shape how these populations fight viruses, or their risk for autoimmune disease" [1].
In other words, the same Denisovan variants that once helped human ancestors survive unfamiliar pathogens in Near Oceania may still be contributing, today, to the autoimmune disease burden in those same populations. The trade reads more like a long, negotiated truce between an ancient adaptation and a modern environment, hashed out in clinics that have only just begun to measure it.
Skeletons, signals, and what the findings change
The immune pathway is not the only place where Denisovan DNA is still doing work. The study also flagged adaptive variants in a gene called TRPS1, which is involved in skeletal development [1][2]. The interesting twist is that TRPS1 is not uniquely Denisovan. The same gene has been independently under positive selection in central African rainforest hunter-gatherers and in highland populations in Ecuador, two very different human groups that have nothing to do with Denisovans [1][2]. That kind of pattern, where unrelated populations land on the same gene for different reasons, is what geneticists call convergent adaptation. It suggests that TRPS1 sits at a sensitive node in human skeletal development, one that has been retuned many times, by many lineages, to meet very different environmental demands. The Denisovan version of the story is one chapter. The African and South American chapters are independent echoes of the same underlying problem: how to build a skeleton well, in very different places. The fact that Denisovan DNA still contributes a working version of the variant in Oceanian populations is, again, evidence that the inheritance is functionally alive rather than fossil.
The work is, on one level, an evolutionary story. On another, it is a precision-medicine story. The same Denisovan variants that helped Pacific ancestors resist ancient pathogens are exactly the variants that today sit behind differences in drug response, infection severity, and autoimmune risk across populations [1][4]. If those variants are not represented in genomic reference panels, the resulting medical tools, from polygenic risk scores to treatment guidelines, will quietly underperform for the people who carry them.
Tucci is blunt about the stakes. "While Denisovans vanished from the Earth thousands of years ago, this research proves that our histories remain deeply intertwined," she said [1]. That intertwining is now, finally, on the map, and on the cell, at the level of working switches that turn human genes up and down.
These are research findings about population-level genetic associations; they are not a basis for individual medical decisions and should not be read as such.
For a field that has spent much of the past decade mapping the archaic roots threaded through the modern genome, the next question is harder. The harder question is what those stretches are doing, today, in living cells, in living people, in clinics that have not yet been built. The 177 genomes from the Bismarcks and the Solomons are a first step. The lab notebooks, in New Haven and beyond, are just opening. If a new medicine is going to work for the people on that windward beach, the work in those notebooks may matter as much as any pill. Could the next Denisovan discovery come, not from a frozen cave in Siberia, but from a freezer in New Haven, opened by a graduate student who has never seen a Denisovan bone?