A label on a reference population describes a genetic profile, not a certified ethnicity
Ancient DNA calculators are built from named reference populations, "Iberian Roman", "Balkan Iron Age", "Anatolian Bronze Age" and so on. It is easy to read a name like that as a precise ethnic or national label, but that is almost never what it actually means, and conflating the two is one of the most common sources of confusion (and, sometimes, genuine controversy) in this hobby.
A real example: what a "Roman-era Anatolian" cluster actually contains
We recently reviewed the underlying sample composition behind a reference group used on a widely-shared calculator, labelled with a name implying Greek ethnic origin. Once we actually counted the individual archaeological samples behind that label, the picture was very different from what the name suggested: only about half of the individuals were recovered at sites in Turkey or Greece. The rest came from Roman-era burial sites scattered right across the empire, Italy, the Balkans, Austria, the Netherlands, even North Africa, with Italy alone contributing more individuals than any single Anatolian site.
None of that makes the underlying genetic data wrong. It genuinely reflects a real, well-documented historical phenomenon: people carrying an East Mediterranean-Anatolian genetic profile were buried at Roman sites throughout the empire, soldiers, traders, freedmen, migrants. What was misleading was the label. Calling that group "Greek" implies a specific ethnic origin for a genetic signature that is really a broader Roman-era Eastern Mediterranean profile, one that a person of no Greek ancestry at all could easily carry if their family history intersected the same migration and trade networks.
Why this matters for how you read your own results
If your report or calculator shows a strong match to a population with an ethnic-sounding name, it is worth asking what actually defines that cluster: is it a tight, geographically and chronologically specific sample set, or a broad composite drawing from many regions that happens to carry a shared genetic signature? The second kind is common and still useful, but the honest way to read it is "your DNA is closest to a profile that shows up in this historical context," not "you are certified as belonging to this modern ethnic group."
This is exactly why, across our own reports, we favour neutral, geography-and-period based names (Roman-provincial Dalmatia, Iron Age Thracian, Bronze Age Balkan core) over ethnic labels wherever the underlying sample set is broader than a single modern identity, and why we keep the individual reference samples in each panel documented and open to review rather than treating the label as the last word.
The takeaway
A confident-sounding population name is a starting point for a question, not proof of an answer. When a result surprises you or does not match your known family history, the most useful next step is not to argue about the label, it is to ask what samples actually sit behind it. That is usually where the real story is.