deCODE genetics

Based on Wikipedia: deCODE genetics

Iceland has roughly the population of a mid-sized American suburb. Three hundred thousand people live on a volcanic island in the North Atlantic, most of them in or around Reykjavík. And yet this tiny nation has become one of the most important places on Earth for understanding human genetics—not despite its small size, but because of it.

The reason comes down to a simple fact: Icelanders know who their ancestors are. Not in the vague way that most of us know our grandparents' names, but in a comprehensive, documented, often centuries-deep way that traces back to the Vikings who first settled the island in the ninth century. When you combine that genealogical richness with modern genomics, something remarkable happens. You can decode not just what's in a person's DNA, but where it came from and what it means.

This is the story of deCODE genetics, a company that bet everything on that insight.

The Harvard Professor Who Went Home

In 1996, Kári Stefánsson was a tenured professor at Harvard Medical School. He had made it. In the ruthless competition of academic medicine, he had secured one of the most coveted positions in the world—a job for life at one of the most prestigious institutions on the planet.

He walked away from it.

Stefánsson returned to Iceland with twelve million dollars in American venture capital and an idea that most of his colleagues considered somewhere between ambitious and delusional. He wanted to use the entire population of his home country as a genetics laboratory.

The timing seemed terrible. In 1996, scientists were just beginning to crack the code of rare genetic diseases—conditions caused by mutations in single genes that could be tracked through small family trees. But Stefánsson wasn't interested in rare diseases. He wanted to understand the common ones: heart disease, diabetes, cancer. The killers that account for most human suffering and death.

The conventional wisdom said this was impossible. These diseases were too complicated. They involved too many genes interacting with too many environmental factors—diet, exercise, smoking, stress. How could you ever tease apart the genetic signal from all that noise?

Stefánsson had an answer, though it sounded crazy at the time: scale. You couldn't find subtle genetic effects by studying a few hundred people. You needed tens of thousands. Eventually hundreds of thousands. You needed, in essence, an entire country.

Why Iceland?

To understand why Stefánsson chose Iceland, you need to understand what makes it genetically unusual.

The island was settled primarily by Norse Vikings starting around 870 AD, with some Celtic admixture from Ireland and Scotland. Then, for about a thousand years, very few people came or went. The population remained isolated, intermarrying within itself generation after generation. This created what geneticists call a "founder population"—a group descended from a relatively small number of original settlers.

Such populations are genetic goldmines. Because everyone shares common ancestors, the genetic background noise is lower. Variants that affect disease risk stand out more clearly against this uniform backdrop. It's like trying to spot a specific bird in a forest versus trying to spot it in a zoo—the more homogeneous the environment, the easier the target becomes visible.

But isolation alone doesn't make Iceland special. What makes it unique is the record-keeping.

Icelanders have been documenting their family connections for over a millennium. The medieval sagas—those epic tales of feuds, voyages, and legendary heroes—are full of genealogical information. The first nominal census anywhere in the world was conducted in Iceland in 1703, recording every person by name. Church parish records extend back centuries. And modern Icelanders, perhaps because there are so few of them, maintain an active interest in who is related to whom.

deCODE partnered with a local software company to digitize all of this. They created a database called Íslendingabók—the Book of Icelanders—that links together virtually every Icelander who has ever lived. By the early 2000s, they could trace the ancestry of any living citizen back to 1703 with near-complete accuracy, and often much further.

This genealogical database became one of deCODE's secret weapons. When you know exactly how thousands of research participants are related to each other, you can do things with genetic data that are impossible otherwise.

The Participation Problem

Having a theoretically ideal population for genetic research is one thing. Actually getting people to participate is another.

This is where deCODE's numbers become almost unbelievable. From its earliest days, more than ninety percent of Icelanders who were asked to participate in the company's research said yes. Not ninety percent of some self-selected group of volunteers—ninety percent of randomly selected citizens approached about taking part.

Participation isn't trivial. It means going to a data collection center, having blood drawn, answering questionnaires, and sometimes undergoing clinical examinations. People do this for no direct personal benefit, on the promise that their contribution might someday help someone else.

By 2003, over one hundred thousand Icelanders had volunteered for one or more of deCODE's research programs. By 2018, that number exceeded one hundred sixty thousand. In a country of three hundred thousand people, this means that more than two-thirds of all adults have donated their DNA and medical information to this single company's research efforts.

Why such extraordinary cooperation? Part of it is surely cultural—Iceland is a small, tight-knit society with strong civic traditions. Part of it may be trust in a company founded and led by a native Icelander rather than foreign scientists. And part of it is probably the intuitive understanding that in a population this small, genetic discoveries might genuinely help people you know.

Whatever the reasons, this level of participation has no parallel anywhere else in the world.

The Database Controversy

Not everyone was enthusiastic about deCODE's ambitions.

In 1998, the company lobbied the Icelandic Parliament to pass the Act on Health Sector Database. This law would have allowed deCODE to create a centralized database containing copies of medical records from across Iceland's national health service. Citizens would be included by default unless they specifically opted out. The company would use this data for commercial research and, in theory, to improve the health system.

The proposal ignited a fierce battle. Critics—including a vocal group of Icelandic activists and numerous international bioethicists—argued that the database represented an unprecedented intrusion into medical privacy. They objected to the opt-out model, which assumed consent rather than requiring it. They worried about a private company having access to an entire nation's health records. And they questioned whether the promised benefits would ever materialize.

The debate played out in local and international media, transforming deCODE from an obscure startup into one of the most controversial enterprises in genomics. In the end, the Iceland Health Sector Database was never built. But the controversy accomplished something else: it forced the world to grapple seriously with the ethical implications of large-scale genetic research.

These questions—about consent, privacy, commercial use of medical data, and the balance between individual rights and collective benefit—would only become more pressing as genetic technology advanced. Iceland was the proving ground where many of these debates first took shape.

The Technology Races

While deCODE was navigating political controversies, it was also racing to prove its scientific approach.

In June 2000, Bill Clinton and Tony Blair jointly announced the completion of the first rough draft of the human genome sequence. This was the culmination of the Human Genome Project—a massive international effort, involving thousands of scientists and billions of dollars, to read out the complete genetic instructions for building a human being.

The achievement was monumental. But it was also, in a sense, just the beginning. Having the sequence of a single reference genome was like having a dictionary—useful for looking things up, but not very informative about how language is actually used. To understand how genetic variation affects human health, you needed to compare many genomes and look for differences that correlated with disease.

This is where deCODE's approach diverged from the mainstream.

The Human Genome Project was essentially an engineering challenge: how do you generate enough raw sequence data to assemble a complete genome? deCODE was tackling a different problem: how do you analyze variation across tens of thousands of genomes to find the genetic basis of disease?

The key insight was that genomes don't exist in isolation. They are inherited. Every person's genome is a mosaic of segments passed down from ancestors, recombined through generations of reproduction. If you know how people are related—which deCODE did, thanks to the genealogies—you can use that information to understand the genome in ways that would otherwise be impossible.

In 2002, deCODE published a genetic map of the human genome consisting of five thousand markers ordered across all the chromosomes. This map, made possible by the genealogical data, was critical to fixing errors in the public Human Genome Project's assembly. It helped improve the accuracy of the reference genome from ninety-three percent to ninety-nine percent.

A company in Iceland, studying Icelanders, had made a fundamental contribution to one of the most important scientific projects in history.

The GWAS Revolution

The mid-2000s brought a technological breakthrough that would transform human genetics: the genotyping chip.

These chips—small glass slides covered with tiny spots of DNA—could simultaneously measure hundreds of thousands of genetic variants in a single sample. Where earlier methods required laboriously testing one variant at a time, a genotyping chip could survey the entire genome in a single experiment.

This technology enabled a new kind of study called a genome-wide association study, usually abbreviated as GWAS (pronounced "gee-wass"). The concept is straightforward: collect DNA from thousands of people with a disease, collect DNA from thousands of healthy controls, and systematically compare them. Any genetic variants that are more common in the disease group than in the controls might be involved in causing the disease.

GWAS set off a global boom in genetic research. Suddenly, scientists everywhere were scanning genomes for disease associations. And deCODE, with its massive collection of DNA samples, detailed medical records, and comprehensive genealogies, was almost perfectly positioned to dominate this new era.

Since 2003, deCODE has discovered and published hundreds of genetic variants linked to susceptibility to scores of diseases. Heart disease. Stroke. Atrial fibrillation. A dozen different types of cancer. Alzheimer's disease. Schizophrenia and other psychiatric disorders. Type 2 diabetes. The list goes on.

A remarkable paper published in Nature Communications in 2019 quantified deCODE's outsized contribution to the field. Between 2007 and 2017, Icelanders accounted for twelve percent of all participants in all published GWAS studies globally. Think about that: a country with 0.004 percent of the world's population contributed twelve percent of the research participants driving discoveries about human genetics.

The same analysis ranked the most impactful GWAS researchers in the world. Kári Stefánsson—the Harvard professor who had walked away from tenure to start a company in his home country—was ranked first.

The Power of Imputation

One of deCODE's cleverest tricks involves a technique called imputation, which is worth understanding because it illustrates how the genealogies multiply the company's capabilities.

Genotyping chips measure hundreds of thousands of variants, but the human genome contains billions of base pairs. No chip can measure them all. However, genetic variants tend to be inherited together in blocks. If you know someone has variant A at one position, you can often predict with high confidence that they also have variant B at a nearby position, because those variants tend to travel together through generations.

This predictability allows scientists to "impute" unmeasured variants based on measured ones. It's like inferring what words are missing from a partially visible sentence—if you can see enough of the context, you can fill in the gaps.

Now add the genealogies. If you know exactly how thousands of people are related, you can impute genetic variants far more accurately than you could from genotyping data alone. The family relationships tell you which segments of DNA people must have inherited from which ancestors, constraining the possibilities and making predictions more precise.

deCODE took this to an extreme. When whole-genome sequencing became economically feasible—reading not just selected variants but the entire three billion base pairs of DNA—the company directly sequenced several thousand Icelanders. Then, using the genealogies as a scaffold, they imputed whole-genome sequence data for virtually the entire population.

In effect, they generated complete genome sequences for hundreds of thousands of people at a tiny fraction of the cost of actually sequencing them all. The first major results of this analysis were published in a special edition of Nature Genetics in 2015, revealing insights that would have been impossible to obtain any other way.

From Discovery to Drug Development

Finding genetic variants associated with disease is scientifically exciting, but translating those discoveries into treatments is what ultimately matters for patients.

In 2012, deCODE was acquired by Amgen, one of the world's largest biotechnology companies. This might have seemed like the end of an independent Icelandic success story, but the acquisition actually amplified deCODE's impact. As an Amgen subsidiary, the company's discoveries could flow directly into drug development pipelines with the resources to turn scientific insights into actual medicines.

The logic connecting genetics to drug development is compelling. If a genetic variant increases risk for a disease, the gene affected by that variant is probably involved in the disease process. And if you can understand how the gene contributes to disease, you might be able to develop a drug that targets it.

This approach has a famous precedent. In families with extremely high cholesterol and early-onset heart disease, researchers discovered rare mutations in a gene called PCSK9. These mutations caused the gene to be overactive, leading to dangerously elevated cholesterol. The discovery suggested that blocking PCSK9 might lower cholesterol—and indeed, PCSK9 inhibitors are now an important class of cholesterol-fighting drugs.

deCODE's vast dataset enables systematic searches for exactly these kinds of rare, high-impact variants. By studying not just common genetic variations but also rare ones—made possible by whole-genome sequencing and the population's comprehensive genealogies—the company can identify genes that represent promising drug targets.

Amgen's investment in deCODE has helped spur a broader industry trend. Other pharmaceutical and biotechnology companies have followed suit, investing heavily in genomics and precision therapeutics. The small Icelandic company that Kári Stefánsson founded with venture capital in 1996 has helped reshape how the global pharmaceutical industry thinks about drug discovery.

The Book of Icelanders

Perhaps the most charming manifestation of deCODE's work is Íslendingabók—the Book of Icelanders.

In 2003, deCODE launched a public-facing, online version of its genealogy database. Any Icelander with a social security number can request access and then explore their family tree, tracing connections back through centuries of ancestors and seeing exactly how they're related to anyone else in the country.

Within its first month online, more than a third of the population had requested passwords. By 2020, the database had over two hundred thousand registered users—about two-thirds of the entire nation—and contained more than nine hundred thousand linked entries, representing the majority of Icelanders who have ever lived.

On an average day, nearly six thousand people—close to two percent of all citizens—consult the database. Icelanders use it to research their ancestry, to settle family disputes, to satisfy curiosity about famous historical figures they might be descended from, and occasionally to check whether they're too closely related to someone they're dating. (There's actually an app for that, nicknamed the "incest prevention alarm.")

In a country that is essentially one enormous extended family, the Book of Icelanders has become part of daily life. It's also a remarkable example of public engagement with genetics research—a constant, visible reminder of the scientific enterprise that Icelanders have collectively made possible.

The Broader Significance

What deCODE has accomplished in Iceland cannot be easily replicated elsewhere. Few places combine the genealogical records, the genetic homogeneity, the small and cooperative population, and the high-quality healthcare system that make this kind of research possible.

But the Iceland experiment has influenced genetics research worldwide.

First, it proved that population-scale genomics could work. When Stefánsson founded deCODE in 1996, the idea of studying tens of thousands of people's genomes seemed impossibly ambitious. Now it's routine. Massive biobank projects like the UK Biobank (with half a million participants) and All of Us in the United States (aiming for one million) owe something to the proof-of-concept that Iceland provided.

Second, deCODE's work demonstrated the power of combining different types of data. Genomes alone are just sequences of letters. Medical records alone are just lists of diagnoses and treatments. Genealogies alone are just family trees. But integrate all three, and you get something far more powerful than any single dataset could provide. This principle of data integration now underlies much of modern biomedical research.

Third, Iceland forced the world to think seriously about the ethics of genetic research. The Health Sector Database controversy of the late 1990s raised questions about consent, privacy, and commercial use of health data that remain relevant today. Every major genomics initiative now grapples with these issues, and the frameworks they use bear traces of the debates that first erupted in Reykjavík.

What the Genes Have Revealed

After more than two decades of research, what has deCODE actually learned about human genetics and disease?

The findings fall into several categories.

For some conditions, specific genetic variants have emerged as reliable risk markers. A variant in a gene called TCF7L2, discovered by deCODE, substantially increases risk for type 2 diabetes and is now standard in genetic risk models. Similar variants have been identified for heart disease, Alzheimer's disease, and many cancers.

More broadly, the research has confirmed that virtually every common disease has a genetic component—but also that this genetic architecture is almost always complex. There is rarely a single gene that causes heart disease or depression or cancer. Instead, there are typically dozens or hundreds of genetic variants, each contributing a small amount of risk, interacting with each other and with environmental factors in ways that are difficult to predict for any individual.

This complexity is humbling. The early hope of genomics—that identifying disease genes would lead straightforwardly to cures—has proven too optimistic. The genetic maps are much messier than anyone expected.

Yet the maps are also more useful than critics predicted. Polygenic risk scores—calculations that combine the effects of many genetic variants—can identify individuals at substantially elevated risk for conditions like heart disease or breast cancer. This information can guide screening, prevention, and treatment decisions. It's not the genetic crystal ball that science fiction promised, but it's a genuine advance in medical practice.

The Future

deCODE's research continues. With each passing year, more Icelanders are directly sequenced, adding depth to the population's genetic portrait. With each technological advance—from chips to sequencing to ever more sophisticated analytical methods—the company extracts more information from its unparalleled dataset.

The questions being pursued have evolved. Early GWAS studies focused on finding any genetic variants associated with disease, regardless of how small their effects. Now the emphasis has shifted toward understanding the biological mechanisms underlying those associations. Finding a variant is just the first step; understanding what it does in cells and tissues is what enables drug development.

deCODE is also investigating aspects of genetics that go beyond simple inheritance. Epigenetics—chemical modifications to DNA that affect gene activity without changing the underlying sequence—adds another layer of complexity. So does somatic mutation—genetic changes that occur during a person's lifetime, particularly relevant to cancer. The company's infrastructure, built for studying inherited variation, is being adapted to address these newer questions.

Meanwhile, Iceland's model continues to inspire imitators. National genome projects have been launched in countries from the United Kingdom to Estonia to Singapore. Large-scale biobanks are collecting DNA and health data from millions of participants worldwide. The age of population genomics that Kári Stefánsson envisioned in 1996 has arrived.

An Unlikely Genetic Superpower

There is something improbable about the whole story. A volcanic island in the North Atlantic, with fewer people than a medium-sized American city, has become one of the most important places on Earth for understanding the human genome. A neurologist who gave up a Harvard professorship to move home bet his career on ideas that most of his colleagues considered dubious. A private company convinced an entire nation to participate in its research.

None of this was inevitable. Different decisions at various points—about technology, about ethics, about business models—could easily have led to different outcomes. The Health Sector Database controversy might have poisoned public trust. The venture capital might have dried up. The science might not have panned out.

But it did pan out. deCODE genetics has contributed more per capita to human genomic knowledge than any other organization on the planet. It has helped establish population genetics as a powerful approach to understanding disease. It has influenced the pharmaceutical industry's approach to drug discovery. And it has demonstrated—in the most tangible possible way—that small countries can punch far above their weight in science.

For Icelanders, participating in deCODE's research has become something like a civic tradition—a way of contributing to scientific progress while learning more about their own heritage and health. For the rest of the world, Iceland stands as proof that with the right combination of resources, cooperation, and vision, it's possible to decode secrets hidden in our DNA that affect us all.