What Does Your Genetic Distance Score Actually Mean?
When MYDNATLAS says you are "1.96" from German Mennonite, what does that number mean? A clear guide to interpreting distance scores.
What Does Your Genetic Distance Score Actually Mean?
You upload your DNA to MYDNATLAS, run the Closest Populations analysis, and see your top match: German Mennonite: 1.96. But what does 1.96 actually mean? Is that good? Bad? How close is close?
The Mathematics
Genetic distance in MYDNATLAS is Euclidean distance in G25 coordinate space. Your DNA is represented as a point in 25-dimensional space, and the distance to any reference population is simply the straight-line distance between those two points:
distance = sqrt( (PC1_you - PC1_ref)^2 + (PC2_you - PC2_ref)^2 + ... + (PC25_you - PC25_ref)^2 )
Smaller distances mean greater genetic similarity.
The Scale
Here's a practical guide to interpreting distance scores:
| Distance | Interpretation | Example |
|---|---|---|
| < 2.0 | Very close match | You likely share significant recent ancestry with this population |
| 2.0 - 3.0 | Close match | Neighboring or historically connected populations |
| 3.0 - 5.0 | Moderate distance | Same broad region but distinct genetic history |
| 5.0 - 10.0 | Distant | Different regions of the same continent |
| 10.0 - 20.0 | Very distant | Different continents or major genetic clusters |
| > 20.0 | Extremely distant | Populations with minimal shared recent ancestry |
Real-World Examples
To make this concrete, here are typical distances between well-known populations:
- English ↔ Scottish: ~1.5 (very close, recent common ancestry)
- English ↔ French: ~2.5 (close neighbors with shared Celtic/Germanic/Roman heritage)
- English ↔ Greek: ~5.0 (same continent, different genetic history)
- English ↔ Japanese: ~18.0 (different continents, minimal overlap)
- English ↔ Yoruba: ~25.0 (maximum Eurasian-African divergence)
Why "Closest" Matters
Your closest population match tells you which reference group your DNA most closely resembles as a whole. This doesn't mean you're "from" that place, it means the combination of ancestral components in your DNA most closely matches the average genetic profile of that population.
For example, a Palestinian person might show:
- Palestinian: 1.5
- Jordanian: 1.8
- Lebanese_Muslim: 2.1
- Syrian: 2.3
- Druze: 2.8
This tells us their DNA clusters tightly with Levantine populations, with gradually increasing distance as we move to more geographically distant groups.
Multi-Way Fits: Going Deeper
While closest population gives you a single-population match, Multi-Way Fits finds the best combination of populations that models your DNA. You might be:
- 60% Palestinian + 40% Anatolian Greek (distance: 0.8)
- 50% Jordanian + 30% Egyptian + 20% Turkish (distance: 0.6)
The lower the fit distance, the better the model explains your actual genetic profile.
Tips for Interpretation
- Look at the top 5-10 matches, not just #1, the cluster pattern reveals more than any single result
- Distance is relative, a score of 3.0 for a European is typical; for a Sub-Saharan African comparing to Eurasian populations, scores above 20 are expected
- Ancient matches have higher distances, ancient populations had less admixture, so modern people rarely match them below 3.0
- Fit error in Multi-Way is different, a Multi-Way fit distance of 2.0 is poor; below 1.0 is excellent
Explore your own distances on MYDNATLAS, upload your DNA and see where you land in the global genetic landscape.
Written by
MYDNATLAS Research Team
MYDNATLAS Team
