Scientists at deCODE Genetics in Reykjavík have built the first Icelandic pangenome, creating a more detailed map of genetic variation in Iceland and uncovering a previously hidden variant associated with early-onset Parkinson’s disease.
The research, published in Nature, found millions of genetic variants that conventional genome mapping had missed.
What’s the Story?
- Researchers at deCODE Genetics created the first Icelandic pangenome, using 698 Icelandic chromosome sequences to build a more representative map of genetic variation.
- When the new reference was used to analyse whole-genome data from 57,630 Icelanders, researchers identified 98.96 million genetic variants — 6.17% more than with conventional genome mapping.
- Among the previously obscured variants was one in GBA1 associated with early-onset Parkinson’s disease. Researchers subsequently replicated the association using genetic data from 429,193 British and Irish participants in UK Biobank.
A Better Map of the Human Genome
The Parkinson’s finding is striking, but it is not really the biggest discovery in the paper.
The more fundamental development is that researchers were able to find the variant because they changed the map they were using to look for it.
Much of modern genetic research relies on a reference genome: a standard DNA sequence against which an individual’s genome can be compared. The approach has transformed genetics, but it comes with a fundamental limitation.
There is no single human genome.
People carry enormous amounts of genetic variation, and comparing everybody against one linear reference can make some regions of an individual’s DNA difficult to map accurately. Geneticists call this reference bias.
A pangenome approaches the problem differently. Instead of representing the human genome as one sequence, it incorporates multiple alternative sequences and the relationships between them.
Think of the difference between a road and a road map.
If every journey has to be described in relation to one prescribed road, alternative routes can be difficult to represent. A map containing many possible routes gives a much better picture of how people actually get from one place to another.
The same principle applies to DNA.
For the new study, researchers assembled 698 Icelandic haplotypes — chromosome sequences inherited from one parent — and incorporated them into the existing Human Pangenome Reference Consortium pangenome.
The resulting reference, called HPRC-ICE, contains 51.41 million small genetic variants.
The researchers also developed two computational methods to make the approach possible at scale. Emblask was developed to construct haplotype-resolved genome assemblies, while Weaver allows short DNA reads to be mapped against pangenomes containing hundreds of haplotypes.
They then put their new map to work.
Whole-genome sequencing data from 57,630 Icelanders produced 98.96 million detected variants, 6.17% more than researchers identified when mapping the same data against a conventional linear reference genome.
The importance is not simply that the number became larger.
The pangenome recovered variation in parts of the genome where conventional mapping performs poorly — variation that the researchers describe as having previously been “camouflaged” by reference bias.
One of those difficult regions contained GBA1.
A Parkinson’s Variant Comes Into View
GBA1 was already known to be important in Parkinson’s disease. What the Icelandic pangenome revealed was a pathogenic variant that conventional mapping had largely failed to detect.
According to deCODE Genetics, Icelandic carriers of the variant were more than seven times as likely to develop Parkinson’s disease before the age of 60 as non-carriers.
The researchers then tested whether the association could be seen outside Iceland.
They carried out targeted remapping using genetic data from 429,193 British and Irish participants in UK Biobank and replicated the association between the GBA1 variant and early-onset Parkinson’s disease.
That distinction is important.
The researchers have not discovered an Icelandic cause of Parkinson’s disease.
Instead, researchers in Iceland developed a better way of seeing genetic variation, identified a medically significant variant that conventional mapping had difficulty detecting, and then found the same disease association in a much larger population outside Iceland.
Nor was Parkinson’s the only example.
The pangenome also revealed a previously obscured variant in CBS, a gene associated with homocystinuria, a rare inherited disorder affecting the body’s processing of certain amino acids. The paper identifies the CBS variant as pathogenic for the disease.
Both discoveries illustrate why the additional variants recovered through pangenome mapping can matter.
Why Iceland?
Iceland has occupied an unusual position in human genetics for decades.
Its comparatively small population, extensive genealogical records and large collections of genetic and medical data have allowed researchers to study relationships between inherited variation and disease at a scale difficult to reproduce elsewhere.
deCODE has previously sequenced thousands of Icelandic genomes and combined genetic information with genealogy and medical records to investigate inherited disease.
The new research changes something even more fundamental: the reference against which those genomes are read.
Rather than treating one DNA sequence as the standard human genome and describing everybody else’s DNA in relation to it, the pangenome incorporates genetic diversity into the reference itself.
The Icelandic contribution is substantial. The researchers created high-quality assemblies from 361 Icelanders, producing the 698 Icelandic haplotypes incorporated into HPRC-ICE.
The resulting pangenome contains 788 haplotypes in total, combining the new Icelandic sequences with sequences from the international Human Pangenome Reference Consortium and existing reference genomes.
That richer map allowed the researchers to recover variants that had effectively been hidden by the limitations of the older reference.
What Does It Mean for Parkinson’s?
For patients, the finding should not be mistaken for an immediate new treatment or diagnostic test.
A genetic association can help researchers understand disease risk and biology, but considerable work separates the identification of a variant from a new therapy.
Nor does finding 6.17% more variants mean scientists suddenly understand 6.17% more disease.
What the study demonstrates is more basic — and potentially more important.
The map scientists use to read the genome affects what they are able to see.
In this case, changing that map revealed a pathogenic Parkinson’s-associated variant that conventional genome mapping had struggled to detect.
And because researchers were able to replicate the association in hundreds of thousands of people outside Iceland, the result extends beyond the particular genetics of the Icelandic population.
The first Icelandic pangenome is therefore both a new resource for studying Icelanders and a demonstration of what human genetics can uncover when a single reference genome is no longer treated as the only road through our DNA.
Read more stories covering Icelandic society, politics, business, and culture at Iceland Review News.
