With what ancient ethnicity do you most identify, and what has DNA told you ?

Thanks for the link to Maciamo's article! I printed it out. It will make a nice addition to my genealogy files.
 
I would have to say Norse and celtic. My autosomal studies say I am genetically closest to Icelanders but I have pure Croatian ancestry.:confused:
 
I would have to say Norse and celtic. My autosomal studies say I am genetically closest to Icelanders but I have pure Croatian ancestry.:confused:

But did the autosomal comparison include Croatian, Slovenian, Bosnian, Serbian or Austrian samples ? I suppose that you are referring to the Genetic Atlas comparison on deCODEme, but it doesn't have any samples from the Balkans or Central Europe. It doesn't mean much to be closer to the Icelandic when the alternative is Italian, French, Basque or Orcadian.
 
Roman, obviously.

Yet Spaniards aren't that much of Roman descent. The main ancestry is Celtiberian, which is the combination of Paleolithic Iberian (Cro-Magnon I2), Neolithic Iberian from the Near East on the southern coasts (E-M78, J2), and Celtic (especially in the centre and north, R1b-P312 and G2a3a). The Phoenicians, Greeks, Romans, Visigoths, Suebi, Vandals, and Arabs had only a minimal (but accumulative) impact on the Spanish genetic pool. There has never been a major migration of people from Italy to Spain, less so than from Germany, the Levant or North Africa to Spain.
 
Maciamo, it's not a matter of genetic markers. It's culture. I have to do much more with a roman than with a celtiberian, iberian....

If there's an ancient culture that fits better than others in western Europe, that's Rome. Even a current germanic-speaker would feel more comfortable in it.
 
Yet Spaniards aren't that much of Roman descent. The main ancestry is Celtiberian, which is the combination of Paleolithic Iberian (Cro-Magnon I2), Neolithic Iberian from the Near East on the southern coasts (E-M78, J2), and Celtic (especially in the centre and north, R1b-P312 and G2a3a). The Phoenicians, Greeks, Romans, Visigoths, Suebi, Vandals, and Arabs had only a minimal (but accumulative) impact on the Spanish genetic pool. There has never been a major migration of people from Italy to Spain, less so than from Germany, the Levant or North Africa to Spain.
That's correct, but I would change Arabs for Berbers, and also the P312 is after the basques most strong in Catalonia where the Iberian substratum is more predominant (M167)
 
But did the autosomal comparison include Croatian, Slovenian, Bosnian, Serbian or Austrian samples ? I suppose that you are referring to the Genetic Atlas comparison on deCODEme, but it doesn't have any samples from the Balkans or Central Europe. It doesn't mean much to be closer to the Icelandic when the alternative is Italian, French, Basque or Orcadian.


Autosomal results can predict which nation you're related to the closest genetically? If so what do they include as examples and which sites offer the analysis?
 
But did the autosomal comparison include Croatian, Slovenian, Bosnian, Serbian or Austrian samples ? I suppose that you are referring to the Genetic Atlas comparison on deCODEme, but it doesn't have any samples from the Balkans or Central Europe. It doesn't mean much to be closer to the Icelandic when the alternative is Italian, French, Basque or Orcadian.

Not true.

There are many public samples now from all over Europe including all the countries you have listed. You are obviously not a member of DecodeMe otherwise you would have known this fact. Regardless, I still match Icelanders at all levels of autosomal DNA ie. at 1Kb, 250 Kb and 5Mb. In fact I match Icelanders better than they match each other most of the time.
 
Last edited:
Autosomal results can predict which nation you're related to the closest genetically? If so what do they include as examples and which sites offer the analysis?

Yes they can. DecodeMe have included reference populations around the world.
 
Not true.

There are many public samples now from all over Europe including all the countries you have listed. You are obviously not a member of DecodeMe otherwise you would have known this fact. Regardless, I still match Icelanders at all levels of autosomal DNA ie. at 1Kb, 250 Kb and 5Mb. In fact I match Icelanders better than they match each other most of the time.

I don't know what game you are playing. DeCODEme only has 9 European populations from 6 countries in its Genetic Atlas : Iceland, Orkney, France, Basque, Italy (unspecified Italian, Tuscan and Sardinian), and Russia (Russian and Adygey). The exact same population are used in their Map of kinship.

So I am not sure where you found your "many public samples from all over Europe". Don't you mean one-to-one comparisons with other customers in the Your Friends section ? That's also possible at 23andMe.

There are 3 main reasons why such comparisons are unreliable :

1) Single individual matches are unrepresentative of a whole country. You will sometimes find good matches in one country, but others that are only distant matches. If your 3 best and 3 worst matches are from the same country they even each other out. There is no option at present to make averages for all members from a selected country. The problem has a solution; it's just not available to us yet.

2) One's best matches vary according to the resolution chosen (fragment size). One's matches at 1Mb aren't the same as at 250Kb or 5Mb. Depending the what you choose your top matches will be in different countries. Some of my best matches at 250Kb (64%) share 0% at 5MB. One person who shares 0.5% at 5MB has only 62.6% at 250Kb. There are so many resolution levels that no consistent ranking can be made. In other words nobody fits neatly into one region.

3) Interpersonal differences within Europe aren't significant. You are unlikely to find any match at 10Mb or 20Mb unless you are related within a few generations. At 5Mb pretty much everyone shares between 0% and 0.5%. At 1Mb it varies between 18% and 20.5%. At 250Kb it ranges from 62.5 to 64%. Differences are tiny and random. What's the significance of 0.1% more or less in common with someone ?
 
Last edited:
I don't know what game you are playing. DeCODEme only has 9 European populations from 6 countries in its Genetic Atlas : Iceland, Orkney, France, Basque, Italy (unspecified Italian, Tuscan and Sardinian), and Russia (Russian and Adygey). The exact same population are used in their Map of kinship.
So I am not sure where you found your "many public samples from all over Europe". Don't you mean one-to-one comparisons with other customers in the Your Friends section ? That's also possible at 23andMe.
There are 3 main reasons why such comparisons are unreliable :
1) Single individual matches are unrepresentative of a whole country. You will sometimes find good matches in one country, but others that are only distant matches. If your 3 best and 3 worst matches are from the same country they even each other out. There is no option at present to make averages for all members from a selected country. The problem has a solution; it's just not available to us yet.
2) One's best matches vary according to the resolution chosen (fragment size). One's matches at 1Mb aren't the same as at 250Kb or 5Mb. Depending the what you choose your top matches will be in different countries. Some of my best matches at 250Kb (64%) share 0% at 5MB. One person who shares 0.5% at 5MB has only 62.6% at 250Kb. There are so many resolution levels that no consistent ranking can be made. In other words nobody fits neatly into one region.
3) Interpersonal differences within Europe aren't significant. You are unlikely to find any match at 10Mb or 20Mb unless you are related within a few generations. At 5Mb pretty much everyone shares between 0% and 0.5%. At 1Mb it varies between 18% and 20.5%. At 250Kb it ranges from 62.5 to 64%. Differences are tiny and random. What's the significance of 0.1% more or less in common with someone ?

Absolutely spot-on. There are autosomal DNA mapping studies completed over the last two years that have shown some success in identifying a number of populations regionally, but things are far from 100%.
 
Last edited:
I don't know what game you are playing. DeCODEme only has 9 European populations from 6 countries in its Genetic Atlas : Iceland, Orkney, France, Basque, Italy (unspecified Italian, Tuscan and Sardinian), and Russia (Russian and Adygey). The exact same population are used in their Map of kinship.

So I am not sure where you found your "many public samples from all over Europe". Don't you mean one-to-one comparisons with other customers in the Your Friends section ? That's also possible at 23andMe.

There are 3 main reasons why such comparisons are unreliable :

1) Single individual matches are unrepresentative of a whole country. You will sometimes find good matches in one country, but others that are only distant matches. If your 3 best and 3 worst matches are from the same country they even each other out. There is no option at present to make averages for all members from a selected country. The problem has a solution; it's just not available to us yet.

2) One's best matches vary according to the resolution chosen (fragment size). One's matches at 1Mb aren't the same as at 250Kb or 5Mb. Depending the what you choose your top matches will be in different countries. Some of my best matches at 250Kb (64%) share 0% at 5MB. One person who shares 0.5% at 5MB has only 62.6% at 250Kb. There are so many resolution levels that no consistent ranking can be made. In other words nobody fits neatly into one region.

3) Interpersonal differences within Europe aren't significant. You are unlikely to find any match at 10Mb or 20Mb unless you are related within a few generations. At 5Mb pretty much everyone shares between 0% and 0.5%. At 1Mb it varies between 18% and 20.5%. At 250Kb it ranges from 62.5 to 64%. Differences are tiny and random. What's the significance of 0.1% more or less in common with someone ?

I am not playing any games. I simply responded to the question in the thread you had created. It was you who sparked a debate with a retort that suggested that my autosomal results through Decodeme were subject to interpretation.

I am simply stating my results and suggesting that there should be some ancient link between my genetic makeup and that of other Icelandic individuals. What is your problem with that?

Your first point (Single individual matches are unrepresentative of a whole country). The genetic atlas averages out your 540,000 SNP matches against large population references from countries around the world. My No. 1 match is Iceland out of over 50 countries around the world. I know there are no central European references, this is where you can compare yourself to your friends that accept your invitations to share data. What you failed to realize is that there is a public reference project underway where participants can contribute their data to forming other reference populations. That is what I was referring to.

Your second point (One's best matches vary according to the resolution chosen (fragment size). No kidding! In fact to be precise your low level (250Kb) matches usually indicate that you are closely related to that gene pool. Higher fragment sizes indicate that you may be related to an individual but not necessarily from that gene pool. Hence the differences found.

Your last comment "Interpersonal differences within Europe aren't significant". Ofcourse they are!:useless:

My highest matches are: 84.70% @ 1Kb Iceland, 64.6% @250Kb Icelander 21.7% @ 1Mb Icelander, 4.6% @3Mb Icelander, 1.1% @ 5Mb Icelander. I think even "blind Freddie" could see some sort of trend there?

I also have a 23andme Finnish RF cousin that I share 0.8% @ 5Mb. Many other participants have also correlated percentage sharing @ 5Mb and predicted RF cousins through 23andme. I have seen values that vary from 0.6% to 2.0%.
 
Given that I'm half Jooo half Brit, I identify both as one of Aharon's descendants (J1c3d Kohen) and Germanic matrilineal (J1c2).:D
 
My highest matches are: 84.70% @ 1Kb Iceland, 64.6% @250Kb Icelander 21.7% @ 1Mb Icelander, 4.6% @3Mb Icelander, 1.1% @ 5Mb Icelander. I think even "blind Freddie" could see some sort of trend there?

I also have a 23andme Finnish RF cousin that I share 0.8% @ 5Mb. Many other participants have also correlated percentage sharing @ 5Mb and predicted RF cousins through 23andme. I have seen values that vary from 0.6% to 2.0%.

How did you compare results at 1Kb ?

Are you sure you were not adopted ? How can a pure Croatian get such high percentage of similarity with Nordics ? I don't have any match as high as that. At 1MB my matches in Europe are between 17% and 20%. I only have 6 matches above 20% (1 Dutch and 5 Scandinavians), but even my highest at 1MB (21.4%) is only 63.9% at 250Kb and 0.6% at 5MB.
 
How did you compare results at 1Kb ?

Are you sure you were not adopted ? How can a pure Croatian get such high percentage of similarity with Nordics ? I don't have any match as high as that. At 1MB my matches in Europe are between 17% and 20%. I only have 6 matches above 20% (1 Dutch and 5 Scandinavians), but even my highest at 1MB (21.4%) is only 63.9% at 250Kb and 0.6% at 5MB.


The global similarity is done at 1kb. If you go to the European global similarity map these are done at 1Kb.

LOL no, I am not adopted. I have really have no idea but I am trying to work it out. This is obviously something ancient (not recent), perhaps Viking? Strange thing is that I have a nordid phenotype as well. I know of another half Croat/British who also has a very high matches with Icelandic. More samples need to be assessed to see if this is a general trend or just something to do with my family.
 
I am Irish so obviously I identify most with Irish ethnicity and would view Scotland as being the closest country culturally to us. Maybe thats because physically it is the closest country to us!!
 
The global similarity is done at 1kb. If you go to the European global similarity map these are done at 1Kb.

Yes, but how do you see it for people in the Friends section ? Or do you mean that the default Icelander is your best match among the 200+ individuals available for comparison as "public friends" ?

LOL no, I am not adopted. I have really have no idea but I am trying to work it out. This is obviously something ancient (not recent), perhaps Viking? Strange thing is that I have a nordid phenotype as well. I know of another half Croat/British who also has a very high matches with Icelandic. More samples need to be assessed to see if this is a general trend or just something to do with my family.

What is strange is that you are closer to this/these Icelander(s) than most Scandinavians would be. Two unrelated individuals from a same country would get about 19 or 20% of similarity at 1MB, not 21.7%. How can you be closer to Icelanders than other Scandinavians or perhaps even some Icelanders themselves ?

Even weirder is that you should have a high similarity with a Finn too. Finns are very different genetically from Scandinavians due to their partial Siberian ancestry and very limited Indo-European admixture (R1a and R1b). All my close matches are Scandinavians or Dutch, but none are Finns. I am closer to Italians and Spaniards than to Finns.
 
I also feel a certain connection to Slavs given that J1c introduced Balto-Slavic languages along with W (mtDNA of course).
 

This thread has been viewed 294065 times.

Back
Top