MyHeritage DNA Test: Why 93% Iberian Doesn't Fit

MyHeritage DNA: 93% Iberian and French ancestors that don't fit

A DNA test promising to reveal your ancestry delivers 93% Iberian and, alongside, a list mentioning France, Belgium, and Corsica. You pay for certainties and receive a table the buyer themselves cannot interpret. This is the paradox underlying MyHeritage's ethnic estimation analysis, a 49€ product (including shipping) that drops to 29€ on promotion. Those who have documented French surnames in their family tree until the mid-17th century are shocked that the breakdown mixes Gallic and Iberian elements without explaining why. And the problem starts there: the saliva sample comes with no instruction manual.

Why does 93% Iberian include France and Corsica?

The most repeated explanation is that borders between southwest European populations are administrative, not genetic. Much of the DNA platforms label as Iberian, French, or northern Italian comes from the same Celtic substrate that overlaps depending on the calculator consulted. Hence, a mainly Iberian result carries references to France or Corsica without being a contradiction.

Some argue that many tools use Basque DNA as a reference, pushing the reading of the rest of the peninsula's genetics as partly French or Italian. Others claim this bias reduces a typical Spaniard's Iberian component to figures around 27%. On the opposite end, it is defended that the result is correct and what fails is the customer's expectation: who expected a certificate of purity receives, in reality, a map of mixtures.

GEDmatch and calculators: where the percentage changes owner

The repeated advice is to download the raw data file and upload it to external platforms. MyHeritage focuses only on ancestry, while other services add health variants. Its raw data exceeds 700,000 markers, almost 100,000 more than 23andMe, but lacks SNP markers to extract maternal haplogroup: with it, you only get the paternal one.

From there, calculators enter. Eurogenes K13 and K15 are the most used for Western Europe; K36 breaks down into clearer ethnic components; MDLP K23b is highlighted as best for Southern Europeans. Results change from one to another, and with them, the interpretation. GEDmatch, the classic free tool, has a fruta for being outdated since 2012. The paid alternative, Illustrative DNA, offers g25 coordinates to refine further.

The price of knowing who you are: from MyHeritage to 23andMe

The price range also marks the product's depth. MyHeritage costs 49€ including shipping and drops to 29€ on sale; Ancestry DNA costs 59€ plus 12€ shipping; 23andMe is the most expensive and rarely discounts. Some defend the expense compared to other hobbies —an annual video game, weekend drinks— while others respond that this money buys an estimate, not a truth.

Result updates are another battleground. Ancestry DNA revises its database yearly; MyHeritage, according to thread comments, hasn't touched it since it started. A user with 23andMe results unupdated since 2022 is told their genotyping chip is the old model and that with the new one, additional components would appear.

A Valencian with 27% Iberian: percentages reopening the debate

An illustrative case is the Eurogenes K36 results of a Valencian user: a 27.94% Iberian, 17.27% Italian, 12.43% North Sea, and 9.36% North Atlantic, plus 5.15% Basque. The circulating breakdown, component by component, disconcerts those expecting a high, clean Spanish percentage.

Another participant sees 57 in the British Isles when, according to another forum member, the norm for an Iberian is 53 or 54, and doesn't understand why they get so much British if they have more French, German, or Northern Italian. The reading of these data again divides: some believe the test exaggerates foreign elements and minimizes local ones; others, that it simply reflects centuries of mestizaje the family tree never documented.

Fundamental criticisms: science or well-sold dice?

In the conversation, deep distrust weighs. A professional comparing results from different companies is cited, concluding that if they used homologated methods, they would coincide, and that the lack of coincidence reveals a product lacking rigor. Against this, the usual defense: they are probabilistic estimates, not diagnoses, and their value lies in cross-referencing with genealogy.

If a basic fact like the Iberian percentage oscillates from 27% to 93% depending on where you upload it, the question is no longer where we come from, but how much of what is sold as science is actually a range with a margin of error. Is it worth paying for a result that changes label depending on the calculator reading it?