23andme update

[Dunai]

Yes, it does exist out of thin air. Do you think I'm actually 47% Italian, too? I'm none of the things the report says. I'm mostly Dodecanese. We have a unique admixture that cannot be accurately shown using those samples they are using at 23andMe. Similar admixture might have existed on the coasts of Anatolia before the modern Turkish profile formed, but now we are unique and require our own category and reference population.

But the Dodecanese islands are next to Turkey, so not surprising at all that someone from there would get so much WANA estimation. I would be the first to admit, there needs to be more fine tunening to be done, since some of the regions are way too generalized, like Eastern European, but at least the WANA region is pretty well sub-categorized, I doubt by now they lack the knowledge about that region, otherwise they wouldn't have made so many categories for it.

[AncientGreek]

But the Dodecanese islands are next to Turkey, so not surprising at all that someone from there would get so much WANA estimation. I would be the first to admit, there needs to be more fine tunening to be done, since some of the regions are way too generalized, like Eastern European, but at least the WANA region is pretty well sub-categorized, I doubt by now they lack the knowledge about that region, otherwise they wouldn't have made so many categories for it.

With the current model, 23andMe's report for Dodecanese islanders would be accurate only if Central Anatolian Greeks were the Anatolian reference population (they're not, only Turks are).

In that case,

Target: Greek_Kos
Distance: 1.0824% / 0.01082426
50.8 Greek_Central_Anatolia
49.2 Italian_Apulia

Since they're not, this is what we should be scoring:

Target: Greek_Kos
Distance: 1.5284% / 0.01528433
69.2 Italian_Apulia
30.8 Armenian

Which shows that 23andMe gives currently gives us too much WANA.

But, the test is designed to show us our last 500 years ancestry is it not? I can tell you that most Dodecanese islanders are mostly native to the area in the last 500 years, thus it saying we are Italian, Cypriot or Iranian is wrong. I'm confused why you think it is acceptable for 23andMe to give us terribly inaccurate reports?

[Leto]

I actually score on the low end of the Dodecanese WANA scale. Vast majority score higher. Many in the 60%-70% WANA range after the latest update. But 23andMe shouldn't be used as a way to see the amount of West Asian admixture. G25 is much more suitable than 23andMe for that purpose.

Try Dodecad K12b averages made by Kaspias
https://www.theapricity.com/forum/sh...-K12b-Averages

[Dunai]

With the current model, 23andMe's report for Dodecanese islanders would be accurate only if Central Anatolian Greeks were the Anatolian reference population (they're not, only Turks are).

In that case,

Target: Greek_Kos
Distance: 1.0824% / 0.01082426
50.8 Greek_Central_Anatolia
49.2 Italian_Apulia

Since they're not, this is what we should be scoring:

Target: Greek_Kos
Distance: 1.5284% / 0.01528433
69.2 Italian_Apulia
30.8 Armenian

Which shows that 23andMe gives currently gives us too much WANA.

But, the test is designed to show us our last 500 years ancestry is it not? I can tell you that most Dodecanese islanders are mostly native to the area in the last 500 years, thus it saying we are Italian, Cypriot or Iranian is wrong. I'm confused why you think it is acceptable for 23andMe to give us terribly inaccurate reports?

It's up to you if you decide to accept or not your results, but I'm speaking as someone who already seen how reliable are different genetic test companies, and thus far 23andMe is the best in estimating. Let's see if future MyHeritage estimates can top it.