Matching phenotypes and genetic admixtures
I am really interested to see what people with a high percentage of one admixture look like.
The highest national average of "West European" is 68.5%, which is the average for Irish Dodecad members. But a handful of people listed in the spreadsheet exceed 70%, and one even reaches 75.1% (DOD298). This makes me wonder what is the highest percentage that a single individual could have (without resorting to intentional cross-breeding to recreate a 100% West European individual). If there was someone with over 90% of West European admixture, I infer that he or she would look quite close to the archetypal ancestral population of "West European" (read North-West European, or maybe R1b Proto-Indo-European).
The pure East European type is even more elusive. Lithuanians score the highest (61.5%). Only 4 project members exceed 60%, with a maximum of 67.8% (DOD468).
The Mediterranean admixture doesn't reach more than 55% of any individual's genes, and usually in the isolated Sardinians.
It's not the case of all admixtures. The 'Northeast Asian' reaches 98.5% in the Koryaks, the 'Southeast Asian" approaches 92% in the Dai and Lahu, the Mbuti Pygmies are 98.3% 'Paleo African', North Kannadi are 81.6% 'South Asian'. In all these populations it should be possible to find individuals with 100% of a single admixture element. In others, I expect to find individuals with over 90% of one component (e.g. the Mozabites who are 76.3% 'Northwest African', Georgians who are 72.3% 'West Asian', and Saudis who are 71.5% 'Southwest Asian'). So we can have a good idea of what phenotypes match these genetic components.
It's ironic that it is for the three European components that the maximum percentages are the lowest.
Looking for the origins of the three European populations
So how comes that Europeans are mixed, and who were the three original populations from whom we inherited most of our genome ?
Mediterranean
It's easy to see why no population has kept a high level of Mediterranean genes with all the migrations that took place from all sides the Mediterranean since the Mesolithic. One might wonder if this "Mediterranean" component is even a coherent genetic element, as it is found in such disparate populations as Moroccans, Sardinians, Scandinavians or Iraqi, and pretty much everywhere in between. If it is, it could be the remnant of an aboriginal Paleolithic European population, men carrying Y-haplogroup I lineages, or even an older population whose Y-DNA has all but disappeared (hg F or IJ ?) because their genes were passed only by women. It actually would make sense if the Mediterranean element represented the IJ branch of the Y-chromosomal tree, as it is as common in the Middle East as in Europe. In other words, it would be a very ancient population, perhaps encompassing all the mtDNA subclades of H and V (and perhaps other haplogroups too).
This would explain why the Mediterranean element is so widespread geographically, but also why it reaches such high levels in northern Europe: nearly 25% in the Dutch and Orcadians, 20-22% in the Irish, British and Germans, and around 15% among Scandinavians. If Neolithic farmers reached northern Europe, they were probably more West Asian, and the West Asian element ranges from 4 to 7% in Scandinavia and the British Isles, which fits better with the percentage of haplogroups G2a and J2.
Let's take a concrete example. Belarusians and Poles have considerably more Mediterranean admixture than Lithuanians (14% and 17% against 6.5%, respectively). They also have a bit more West Asian and Southwest Asian (total 2.3% for Belarusians, 3.8% for Poles and 0.5% for Lithuanians). All three have a lot of R1a and some R1b. The main difference is that Belarusians and Poles have considerably more haplogroup I2 and J (20.5% and 13.5% against 7% for Lithuanians) as well as Neolithic haplogroups G2a and E1b1b (total 10.5% and 7.5% against 1.5% for Lithuania). It doesn't match exactly, but there is a trend. Haplogroups G2a and E1b1b are higher than the autosomal DNA from the Middle-East, but that's probably because they were diluted on the way by mixing with other European populations in the Balkans, since it was typically hunter-gatherers who sent their women to marry male farmers rather than the other way round (therefore the Y-DNA remained Middle Eastern, but the mtDNA became European).
North-West European
The West European component of the Dodecad Project being really a North-West European (NWE) one, I will use this denomination here.
Based on my analysis, the NWE component correlates best with haplogroups R1b1b2a1 and I1.
Haplogroup I1 is a special case within haplogroup IJ because it experienced a dramatic bottleneck in the late Neolithic/early Bronze Age that surely had a serious impact on its autosomes. If the re-expansion of I1 with the flourishing of the early Germanic culture happened after the last I1 lineages blended with an R1b1b2a1 dominated population (such as in the Netherlands or Denmark, cradles of the Germanic civilisation), then it makes sense that we should see a correlation between the percentage of I1+R1b and the North-West European component.
In almost every case, we notice an imbalance towards a higher percentage of R1b than actual autosomes (except in Scandinavia and in Romania, where it is matches fairly well).
Ireland has 86% but 68.5% of NWE.
Britain has 81% of R1b+I1 but 66% of NWE.
Germany has 60% of R1b+I1 but 53% of NWE.
France has 70% of R1b+I1 but 52% of NWE.
Spain has 70% of R1b+I1 but 42% of NWE.
Portugal has 58% of R1b+I1 but 40% of NWE.
Why is that so ? I believe it has to do with the quick replacement of native lineages by R1b during the Bronze Age and the genetic predisposition of R1b to father more sons than other haplogroups.
This imbalance is particularly obvious in South Italy and Turkey, where the percentage of R1b is about twice higher than the NWE component. This is simply because a lot of R1b isn't R1b1b2a1, but older Middle Eastern subclades, that are not associated with the the spread of the same people (namely the Proto-Indo-Europeans).
The percentage of mtDNA lineages should balance the difference between autosomal and Y-DNA ratios.
Eastern Europe is the only region where the NWE autosomes exceed R1b+I1, and sometimes to a 3:1 ratio.
Belarus has 13% of R1b+I1 but 28% of NWE
Russia has 11% of R1b+I1 but 32% of NWE
Poland has 23% of R1b+I1 but 35% of NWE
Hungary has 25% of R1b+I1 but 37% of NWE
I believe this is because R1b1b2a1 originated in Eastern Europe, in the Pontic steppes north of the Black Sea, expanded to Central, Northern, Western and Southern Europe, but was ultimately replaced by R1a tribes from northern Russia (the Balto-Slavs) and Central Asia (Scythians, Bulgars, etc.), who killed many of the R1b men left in Eastern Europe, and took their women (the same process as R1b people did when they invaded Europe). This is why the percentage of R1b has diminished in Eastern Europe, but NWE autosomes survived through maternal lineages.
East European
The percentage of East European component matches almost exactly the frequencies for haplogroup R1a. It's truly amazing : 2-3% in Ireland, 5-6% in the Netherlands, 6-7% in Turkey, 12% in Greece, 17% in Germany, 22% in Romania, around 30% in Hungary, around 50% in Poland, Belarus and Russia... Only Scandinavia has a bit more R1a (28% in Norway, 19% in Sweden) than East European autosomes (9.5% and 13%), perhaps due to a founder effect.
The case of Finland
One major exception is Finland, which is a very peculiar case indeed. Finnish people are overwhelmingly European autosomally (93%, against 6% Siberian), and do look Northern European, but their Y-DNA is in great part the Siberian N1c1 (nearly 60%). The best explanation is that these N1c1 lineages were actually for the biggest part autosomally European when they arrived in Finland. As both R1a and R1b are low in Finland (7.5% and 3.5%) but the East European autosomes reach 33% and the North-West European 53%, it means that N1c1 lineages are hiding both East and North-West European autosomes.
It is usually assumed that N1c1 came to Finland from Siberia, but what if it first reach Fennoscandia, then recolonised Finland from northern Sweden, bringing I1, R1a, R1b and Swedish N1c1 in the operation ? The whole group would have been autosomally close to the Swedes, and a founder effect among the ruling class would have spread N1c1 with European autosomes. Sweden itself has only 0.5% of Northeast Asian/Siberian autosomes but 9% of N1c1, mostly concentrated in the north of the country, where an expansion towards Finland would have been most likely. The 6% of Siberian autosomes in Finland would represent the N1c1 that was already in Finland before the recolonisation from northern Sweden.
I am really interested to see what people with a high percentage of one admixture look like.
The highest national average of "West European" is 68.5%, which is the average for Irish Dodecad members. But a handful of people listed in the spreadsheet exceed 70%, and one even reaches 75.1% (DOD298). This makes me wonder what is the highest percentage that a single individual could have (without resorting to intentional cross-breeding to recreate a 100% West European individual). If there was someone with over 90% of West European admixture, I infer that he or she would look quite close to the archetypal ancestral population of "West European" (read North-West European, or maybe R1b Proto-Indo-European).
The pure East European type is even more elusive. Lithuanians score the highest (61.5%). Only 4 project members exceed 60%, with a maximum of 67.8% (DOD468).
The Mediterranean admixture doesn't reach more than 55% of any individual's genes, and usually in the isolated Sardinians.
It's not the case of all admixtures. The 'Northeast Asian' reaches 98.5% in the Koryaks, the 'Southeast Asian" approaches 92% in the Dai and Lahu, the Mbuti Pygmies are 98.3% 'Paleo African', North Kannadi are 81.6% 'South Asian'. In all these populations it should be possible to find individuals with 100% of a single admixture element. In others, I expect to find individuals with over 90% of one component (e.g. the Mozabites who are 76.3% 'Northwest African', Georgians who are 72.3% 'West Asian', and Saudis who are 71.5% 'Southwest Asian'). So we can have a good idea of what phenotypes match these genetic components.
It's ironic that it is for the three European components that the maximum percentages are the lowest.
Looking for the origins of the three European populations
So how comes that Europeans are mixed, and who were the three original populations from whom we inherited most of our genome ?
Mediterranean
It's easy to see why no population has kept a high level of Mediterranean genes with all the migrations that took place from all sides the Mediterranean since the Mesolithic. One might wonder if this "Mediterranean" component is even a coherent genetic element, as it is found in such disparate populations as Moroccans, Sardinians, Scandinavians or Iraqi, and pretty much everywhere in between. If it is, it could be the remnant of an aboriginal Paleolithic European population, men carrying Y-haplogroup I lineages, or even an older population whose Y-DNA has all but disappeared (hg F or IJ ?) because their genes were passed only by women. It actually would make sense if the Mediterranean element represented the IJ branch of the Y-chromosomal tree, as it is as common in the Middle East as in Europe. In other words, it would be a very ancient population, perhaps encompassing all the mtDNA subclades of H and V (and perhaps other haplogroups too).
This would explain why the Mediterranean element is so widespread geographically, but also why it reaches such high levels in northern Europe: nearly 25% in the Dutch and Orcadians, 20-22% in the Irish, British and Germans, and around 15% among Scandinavians. If Neolithic farmers reached northern Europe, they were probably more West Asian, and the West Asian element ranges from 4 to 7% in Scandinavia and the British Isles, which fits better with the percentage of haplogroups G2a and J2.
Let's take a concrete example. Belarusians and Poles have considerably more Mediterranean admixture than Lithuanians (14% and 17% against 6.5%, respectively). They also have a bit more West Asian and Southwest Asian (total 2.3% for Belarusians, 3.8% for Poles and 0.5% for Lithuanians). All three have a lot of R1a and some R1b. The main difference is that Belarusians and Poles have considerably more haplogroup I2 and J (20.5% and 13.5% against 7% for Lithuanians) as well as Neolithic haplogroups G2a and E1b1b (total 10.5% and 7.5% against 1.5% for Lithuania). It doesn't match exactly, but there is a trend. Haplogroups G2a and E1b1b are higher than the autosomal DNA from the Middle-East, but that's probably because they were diluted on the way by mixing with other European populations in the Balkans, since it was typically hunter-gatherers who sent their women to marry male farmers rather than the other way round (therefore the Y-DNA remained Middle Eastern, but the mtDNA became European).
North-West European
The West European component of the Dodecad Project being really a North-West European (NWE) one, I will use this denomination here.
Based on my analysis, the NWE component correlates best with haplogroups R1b1b2a1 and I1.
Haplogroup I1 is a special case within haplogroup IJ because it experienced a dramatic bottleneck in the late Neolithic/early Bronze Age that surely had a serious impact on its autosomes. If the re-expansion of I1 with the flourishing of the early Germanic culture happened after the last I1 lineages blended with an R1b1b2a1 dominated population (such as in the Netherlands or Denmark, cradles of the Germanic civilisation), then it makes sense that we should see a correlation between the percentage of I1+R1b and the North-West European component.
In almost every case, we notice an imbalance towards a higher percentage of R1b than actual autosomes (except in Scandinavia and in Romania, where it is matches fairly well).
Ireland has 86% but 68.5% of NWE.
Britain has 81% of R1b+I1 but 66% of NWE.
Germany has 60% of R1b+I1 but 53% of NWE.
France has 70% of R1b+I1 but 52% of NWE.
Spain has 70% of R1b+I1 but 42% of NWE.
Portugal has 58% of R1b+I1 but 40% of NWE.
Why is that so ? I believe it has to do with the quick replacement of native lineages by R1b during the Bronze Age and the genetic predisposition of R1b to father more sons than other haplogroups.
This imbalance is particularly obvious in South Italy and Turkey, where the percentage of R1b is about twice higher than the NWE component. This is simply because a lot of R1b isn't R1b1b2a1, but older Middle Eastern subclades, that are not associated with the the spread of the same people (namely the Proto-Indo-Europeans).
The percentage of mtDNA lineages should balance the difference between autosomal and Y-DNA ratios.
Eastern Europe is the only region where the NWE autosomes exceed R1b+I1, and sometimes to a 3:1 ratio.
Belarus has 13% of R1b+I1 but 28% of NWE
Russia has 11% of R1b+I1 but 32% of NWE
Poland has 23% of R1b+I1 but 35% of NWE
Hungary has 25% of R1b+I1 but 37% of NWE
I believe this is because R1b1b2a1 originated in Eastern Europe, in the Pontic steppes north of the Black Sea, expanded to Central, Northern, Western and Southern Europe, but was ultimately replaced by R1a tribes from northern Russia (the Balto-Slavs) and Central Asia (Scythians, Bulgars, etc.), who killed many of the R1b men left in Eastern Europe, and took their women (the same process as R1b people did when they invaded Europe). This is why the percentage of R1b has diminished in Eastern Europe, but NWE autosomes survived through maternal lineages.
East European
The percentage of East European component matches almost exactly the frequencies for haplogroup R1a. It's truly amazing : 2-3% in Ireland, 5-6% in the Netherlands, 6-7% in Turkey, 12% in Greece, 17% in Germany, 22% in Romania, around 30% in Hungary, around 50% in Poland, Belarus and Russia... Only Scandinavia has a bit more R1a (28% in Norway, 19% in Sweden) than East European autosomes (9.5% and 13%), perhaps due to a founder effect.
The case of Finland
One major exception is Finland, which is a very peculiar case indeed. Finnish people are overwhelmingly European autosomally (93%, against 6% Siberian), and do look Northern European, but their Y-DNA is in great part the Siberian N1c1 (nearly 60%). The best explanation is that these N1c1 lineages were actually for the biggest part autosomally European when they arrived in Finland. As both R1a and R1b are low in Finland (7.5% and 3.5%) but the East European autosomes reach 33% and the North-West European 53%, it means that N1c1 lineages are hiding both East and North-West European autosomes.
It is usually assumed that N1c1 came to Finland from Siberia, but what if it first reach Fennoscandia, then recolonised Finland from northern Sweden, bringing I1, R1a, R1b and Swedish N1c1 in the operation ? The whole group would have been autosomally close to the Swedes, and a founder effect among the ruling class would have spread N1c1 with European autosomes. Sweden itself has only 0.5% of Northeast Asian/Siberian autosomes but 9% of N1c1, mostly concentrated in the north of the country, where an expansion towards Finland would have been most likely. The 6% of Siberian autosomes in Finland would represent the N1c1 that was already in Finland before the recolonisation from northern Sweden.
Last edited: