Why Your Ancestry Matters for Risk Scores
When you read your Genomisaur health report, your percentile rankings are calculated against a reference population. Which population, and why does it matter? Ancestry is one of the bigger factors in how to interpret a polygenic risk score, so it's worth understanding.
How Reference Populations Work
A polygenic risk score is built from genome-wide association studies (GWAS) — large studies that identify which genetic variants are associated with a condition. Those studies are conducted on specific populations, and the effect sizes they discover reflect the genetics of those populations.
When we calculate your PRS, we sum up the effects of thousands of variants in your genome using the weights from those studies. Your raw score is then compared to a reference panel — a group of people whose scores have been pre-calculated — to produce your percentile.
The Ancestry Gap in Genetics Research
Here's the catch: the vast majority of GWAS to date have been conducted on people of European descent. The effect sizes used in most PRS models are most accurate for European-ancestry individuals and may be less precise for people of other backgrounds.
That doesn't mean your PRS is wrong if you have non-European ancestry. The underlying biology is shared across humans. But the calibration may be looser. A score in the 80th percentile may be slightly over- or under-estimated depending on how well your ancestry matches the study population.
How We Handle It
Our pipeline uses ancestry-aware percentile rankings where possible, comparing your scores against reference panels that better match your genetic background. That produces more meaningful percentiles than a one-size-fits-all approach.
We also flag conditions where the underlying GWAS data is limited for certain ancestries, so you know which results to interpret with extra caution.
What This Means for You
If you have European ancestry, your PRS percentiles are likely accurate. The studies they're based on included millions of participants with similar genetic backgrounds.
If you have African, East Asian, South Asian, or mixed ancestry, your results are still informative but less precisely calibrated. The direction of risk (higher vs. lower) is generally reliable; the exact percentile should be read as approximate.
This is an active area of research and the field is moving fast. Large-scale GWAS in diverse populations are underway, and future PRS models will be more accurate across all ancestries.
The Bigger Picture
Even with imperfect calibration, knowing whether your genetic loading for a condition is high, average, or low is useful for thinking about prevention and screening. We swap in newer, more diverse studies as they become available.