Three distinct genetic components, not two, reshaped the map of Japanese origins
From the islands of Japan, a sweeping genetic study has quietly redrawn the map of human ancestry and its long shadow over living bodies. Researchers at RIKEN, sequencing the full genomes of over three thousand Japanese individuals, have replaced a long-held two-source model of Japanese origins with a tripartite framework — finding distinct ancestral components in Okinawa, the northeast, and the west — while also tracing threads of Neanderthal and Denisovan DNA that still shape disease risk today. The work reminds us that identity, even at the molecular level, is rarely as simple as we once believed, and that the migrations of ancient peoples continue to speak through the health of their descendants.
- A foundational assumption in population genetics — that Japanese ancestry flows from two sources — has been overturned by the largest genomic survey of Japan ever conducted.
- The discovery that two major hereditary breast cancer mutations follow distinct regional ancestry lines creates urgent new questions about how cancer screening is distributed across Japan's geography.
- Ancient DNA inherited from Neanderthals and Denisovans is not merely a curiosity — one segment near a gene targeted by semaglutide correlates with diabetes risk and is absent in Europeans, raising alarms about whether widely used drugs perform equally across populations.
- Nearly 9,800 genetic variants previously invisible to major databases have now been catalogued, expanding the frontier of what clinicians can detect and act upon in East Asian patients.
- The JEWEL dataset is now positioned as a foundational resource, with researchers and medical practitioners looking toward it to build more precise, population-specific approaches to disease prevention and treatment.
Scientists at Japan's RIKEN Center for Integrative Medical Sciences have completed a landmark genetic survey, sequencing the full genomes of 3,256 people across seven regions of Japan. The resulting dataset, called JEWEL, dismantles a model that has guided the field for decades. Rather than two ancestral sources shaping the Japanese people, the study identifies three distinct genetic components — one concentrated in western Japan, bearing strong Han Chinese ancestry; another in the northeast, linked to ancient Korean populations of the Three Kingdoms period and to the Jomon hunter-gatherers; and a third in Okinawa, representing its own separate lineage.
The geographic precision of this tripartite model extends into the realm of disease. Two hereditary breast cancer mutations common in Japan were found to follow ancestral lines with striking clarity — the BRCA1 mutation clustering in people of northeastern ancestry, and BRCA2 in those of western ancestry. The implication is that these mutations arrived separately, carried by different populations at different times, and that cancer screening strategies may need to account for regional genetic heritage.
The JEWEL dataset also catalogued nearly 18,500 loss-of-function variants across more than 9,000 genes, with close to 9,800 of them never before recorded in major genetic databases. Among the findings: a patient with two faulty copies of the ABCC2 gene consistent with a rare liver disorder, and six individuals with damaging variants in PTPRD, three of whom had suffered serious cardiovascular or renal events.
Perhaps the most far-reaching discovery involves ancient inherited DNA. The team identified 44 genetic segments derived from Neanderthals and Denisovans still present in Japanese genomes. One Denisovan segment correlates with type 2 diabetes risk. Another, of Neanderthal origin, sits near GLP1R — the gene targeted by diabetes drugs including semaglutide — and also associates with diabetes. Crucially, this Neanderthal segment appears in East Asian populations but not in Europeans, raising the possibility that common diabetes medications may not perform identically across ancestries.
The researchers also detected strong signals of natural selection in genes governing alcohol metabolism and immune response, with notably weaker signals in Okinawa — a finding they marked for further study. Taken together, the work reveals how ancient migrations and evolutionary pressures continue to shape the health of living people, and positions the JEWEL dataset as a cornerstone for personalized medicine in East Asian populations.
Researchers at Japan's RIKEN Center for Integrative Medical Sciences have completed one of the most comprehensive genetic surveys of the Japanese population to date, sequencing the full genomes of 3,256 people spread across seven regions—from the main islands to Okinawa. The work, published in Science Advances and led by Xiaoxi Liu, upends a model that has dominated the field for years. Where scientists once saw two ancestral sources shaping the Japanese people, this new dataset reveals three distinct genetic components, each concentrated in different parts of the archipelago.
The traditional understanding held that Japanese ancestry flowed from two main sources. The new analysis, built from a dataset the team calls JEWEL, shows the picture is more intricate. Western Japan carries the strongest genetic signature of Han Chinese ancestry. Northeastern Japan tells a different story—its people show closer ties to ancient Korean populations from the Three Kingdoms period and to the Jomon, the hunter-gatherers who inhabited the islands thousands of years before. Okinawa, meanwhile, represents a third distinct ancestral component. This tripartite model, as researchers describe it, suggests that Japan's genetic landscape was shaped by multiple waves of migration and settlement, each leaving its mark in different regions.
What makes this discovery particularly striking is how it illuminates the geography of disease. The team identified two hereditary breast cancer mutations common in Japan and found they follow these ancestral lines with remarkable precision. The BRCA1 mutation appears predominantly in people carrying northeastern ancestry, while BRCA2 concentrates in those with western ancestry. This pattern suggests the mutations arrived separately, carried by different populations at different times. Understanding this geographic distribution could reshape how clinicians screen for cancer risk across Japan's regions.
Beyond ancestry, the JEWEL dataset catalogued 18,481 loss-of-function variants—genetic segments where a gene has been disabled—spread across 9,045 genes. Nearly 9,800 of these variants had never been recorded in major genetic databases before. The team documented one patient carrying two faulty copies of the ABCC2 gene, a finding consistent with Dubin-Johnson syndrome, a rare liver condition marked by chronic jaundice. Six individuals carried damaging variants in the PTPRD gene; three of them had experienced serious health events including heart attacks, kidney failure, and high blood pressure. These discoveries expand the catalog of genetic risk factors available to researchers and clinicians.
Perhaps most intriguingly, the study traced genetic segments inherited from Neanderthals and Denisovans—ancient human species that interbred with modern humans tens of thousands of years ago. The researchers identified 44 such segments still present in Japanese genomes today. One segment near the NKX6-1 gene, inherited from Denisovans, correlates with type 2 diabetes risk. Another, derived from Neanderthals and located near GLP1R—a gene targeted by diabetes medications including semaglutide—also associates with diabetes. This Neanderthal segment appears in East Asian populations but is absent in Europeans, a difference that could have real implications for how diabetes drugs perform across different populations.
The team also detected strong signals of recent natural selection in genes governing alcohol metabolism and immune function, particularly in clusters like ADH and ALDH2, and in the major histocompatibility complex. These signals were noticeably weaker in Okinawa, a finding researchers flagged as warranting closer examination. The work opens a window onto how human populations adapted to their environments over millennia, and how those ancient adaptations continue to shape health and disease today. With nearly 10,000 previously unknown genetic variants now catalogued, the JEWEL dataset promises to inform personalized medicine approaches tailored to East Asian populations for years to come.
Notable Quotes
The tripartite origins model goes beyond the traditional two-ancestry framework— RIKEN researchers, published in Science Advances
The Hearth Conversation Another angle on the story
Why does it matter that Japan has three ancestral sources instead of two?
Because it changes how we understand migration and settlement. Two sources suggested a simpler story. Three sources means different regions have genuinely different genetic histories—and those histories predict where certain diseases show up.
The breast cancer mutations following ancestral lines—is that coincidence?
No. It suggests each mutation arrived with a different population group, centuries or millennia ago. Once you know the ancestry pattern, you can predict who's at higher risk in different regions.
What's the practical use of finding 9,800 new genetic variants?
Right now, mostly research. But each variant is a clue to disease susceptibility. Over time, as we understand what they do, they become tools for screening and prevention.
The Neanderthal and Denisovan DNA—how much of our genome is that?
A small fraction, but concentrated in specific places. What matters is that these ancient segments correlate with modern diseases. And they're distributed differently across populations, which means a drug that works well in Europe might perform differently in East Asia.
Why would Okinawa show weaker signals of selection for alcohol metabolism?
That's the question the researchers are asking. It could mean Okinawa's population history was different—different migrations, different environmental pressures. It deserves follow-up study.