K36 To what do Eurogenes K36 admixtures really correspond?

[Maciamo]

Eurogenes' K36 calculator is quite popular because it has a high number of components, which gives an impression of accuracy. Unfortunately the component names have not been properly labelled and can indeed be very misleading. That is what I am going to try to rectify here. I have run about 20 ancient genomes from the Mesolithic to the early Middle Ages in this calculator to see what component were present in each population, and at what frequency.

One thing I have been able to establish is that some component correspond exclusively to Neolithic farmers, while others were exclusively Mesolithic European or Bronze Age Proto-Indo-European. Here they are:

Mesolithic hunter-gatherer components

- East_Central_Euro
- Eastern_Euro
- Fennoscandian
- North_Atlantic
- North_Sea

These were found in WHG (Spain), SHG (Sweden) and EHG (Yamna, Afanasevo, Corded Ware).

Neolithic farmer components

- Arabian
- Armenian
- East_Balkan
- East_Med
- Iberian
- Italian
- Near_Eastern
- North_African
- Omotic
- West_Med

In Neolithic HUngary (Alföld LBK) and Germany (LBK), the main components were Iberian, Italian and West_Med.
In Funnelbeaker Sweden, had high levels of Iberian, East_Balkan and West_Med, but no Italian at all.

Bronze Age Steppe components

All the same components as in Mesolithic Europeans were found, but also:

- North_Caucasian
- South_Central_Asian
- Volga-Ural

Plus some traces of other admixtures that do not consistently show up, like Amerindian and African.

The "French" component was absent in Yamna, but is found at low frequency (2.5%) in the contemporaneous Afanasievo culture in the Altai (which is otherwise almost identical to Yamna autosomally), at higher frequency in Sintashta (10%) and Corded Ware Denmark (11.5%), and even higher in Corded Ware Poland (26%) and in Nordic Bronze Age Denmark (31.5%), but then sharply decreases in Iron Age Sweden (6%).

The Central_Euro was absent from Mesolithic Spain, Neolithic Germany and Afanasievo and was only at 1% in Yamna. It seems to peak in Central Europe and Scandinavia, but it isn't a Steppe component. It seems to go hand in hand with the French component and could both be associated with Y-haplogroup I1, which was only found in central Europe and Scandinavia before the Germanic migrations in the Iron Age.

The Basque component was found in WHG (8% and 11%), in Funnelbeaker Sweden (2% and 28%), in Chalcolithic Hungary (8%), in Corded Ware Poland (14.5%), Sintashta (1.5%), and in the Nordic Bronze Age (3%). It the component with the least clear pattern.

The 25,000 year-old Mal'ta boy from Siberia (Y-DNA R* and mtDNA U*), who was used to model the ANE admxiture, had East_Central_Euro, Eastern_Euro and Fennoscandian like Mesolithic Europeans, but lacked the Basque, French, Central_Euro, North_Atlantic and North_Sea. Those five latter components are therefore not associated with Y-haplogroup R, but with C1a2, F and I.


Here is the frequency of a few components by prehistoric culture.


Central_Euro (central-north European)

- Corded Ware Poland: 25%
- Corded Ware Denmark: 16%
- Sintashta: 12%
- Corded Ware Germany (Esperstedt): 9%
- Iron Age England : 7.5%
- Iron Age Sweden : 7.5%
- Mesolithic Sweden: 6%
- Yamna: 1%
- Nordic Bronze Age Denmark: 0%
- Afanasievo : 0%
- Mesolithic Hungary: 0%
- Mesolithic Karelia: 0%
- Mesolithic Luxembourg: 0%
- Mesolithic Samara: 0%
- Mesolithic Spain: 0%
- Neolithic Europe: 0%
- Palaeolithic Siberia: 0%

East_Central_Euro (very Mesolithic European, perhaps linked to mtDNA U4 or U5)

- Mesolithic Sweden: 43.5%
- Mesolithic Hungary: 30%
- Mesolithic Samara: 26%
- Corded Ware Poland: 23%
- Mesolithic Spain: 20.5%
- Mesolithic Luxembourg: 20.5%
- Afanasievo : 16.5%
- Sintashta: 15%
- Yamna: 10%
- Mesolithic Karelia: 9%
- Iron Age Sweden : 7.5%
- Nordic Bronze Age Denmark: 6%
- Corded Ware Denmark: 5.5%
- Palaeolithic Siberia: 4.5%
- Iron Age England : 3.5%
- Corded Ware Germany (Esperstedt): 3%
- Neolithic Europe: 0%

Eastern_Euro (Ancient North Asian?)

- Mesolithic Karelia: 23.5%
- Palaeolithic Siberia: 17.5%
- Mesolithic Samara: 17%
- Yamna: 15%
- Corded Ware Germany (Esperstedt): 12.5%
- Sintashta: 12%
- Mesolithic Spain: 11%
- Mesolithic Luxembourg: 10%
- Iron Age Sweden : 7.5%
- Iron Age England : 5%
- Mesolithic Hungary: 3%
- Mesolithic Sweden: 2.5%
- Afanasievo : 0%
- Nordic Bronze Age Denmark: 0%
- Corded Ware Poland: 0%
- Corded Ware Denmark: 0%
- Neolithic Europe: 0%

Fennoscandian (Mesolithic Europe and North Asia, perhaps linked to mt-haplogroup U)

- Mesolithic Luxembourg: 39%
- Mesolithic Samara: 36%
- Mesolithic Spain: 33.5%
- Mesolithic Karelia: 33%
- Mesolithic Hungary: 32.5%
- Palaeolithic Siberia: 21.5%
- Nordic Bronze Age Denmark: 19.5%
- Corded Ware Germany (Esperstedt): 18%
- Iron Age Sweden : 15%
- Afanasievo : 14.5%
- Yamna: 14.5%
- Mesolithic Sweden: 10%
- Iron Age England : 9.5%
- Corded Ware Denmark: 8%
- Sintashta: 7.5%
- Corded Ware Poland: 0%
- Neolithic Europe: 0%

French (especially Baltic; similar to Central_Euro)

- Nordic Bronze Age Denmark: 31.5%
- Corded Ware Poland: 26%
- Corded Ware Denmark: 11.5%
- Sintashta: 10%
- Iron Age England : 9%
- Iron Age Sweden : 6.5%
- Afanasievo : 2.5%
- Corded Ware Germany (Esperstedt): 2%
- Yamna: 0%
- Mesolithic Hungary: 0%
- Mesolithic Karelia: 0%
- Mesolithic Luxembourg: 0%
- Mesolithic Samara: 0%
- Mesolithic Spain: 0%
- Mesolithic Sweden: 0%
- Neolithic Europe: 0%
- Palaeolithic Siberia: 0%

North_Atlantic (mostly northern European)

- Sintashta: 21.5%
- Iron Age Sweden : 19.5%
- Mesolithic Sweden: 19.5%
- Iron Age England : 19%
- Afanasievo : 16.5%
- Corded Ware Denmark: 15.5%
- Corded Ware Germany (Esperstedt): 15%
- Mesolithic Hungary: 10%
- Yamna: 9%
- Mesolithic Luxembourg: 9%
- Mesolithic Spain: 8%
- Mesolithic Karelia: 7%
- Mesolithic Samara: 0%
- Nordic Bronze Age Denmark: 0%
- Corded Ware Poland: 0%
- Neolithic Europe: 0%
- Palaeolithic Siberia: 0%

North_Sea (Mesolithic but especially SHG)

- Corded Ware Denmark: 28%
- Afanasievo : 26%
- Iron Age Sweden : 24.5%
- Corded Ware Germany (Esperstedt): 20%
- Mesolithic Hungary: 19.5%
- Mesolithic Sweden: 18.5%
- Iron Age England : 16.5%
- Mesolithic Spain: 16%
- Sintashta: 15%
- Mesolithic Luxembourg: 13%
- Mesolithic Samara: 11.5%
- Nordic Bronze Age Denmark: 11.5%
- Yamna: 11.5%
- Mesolithic Karelia: 5%
- Palaeolithic Siberia: 0.5%
- Corded Ware Poland: 0 %
- Neolithic Europe: 0%

 

[Twilight]

What an amazing amount of work you put into this Calculator. Thank you for posting. Although question, If you are; let's just say 10% Iberian, does being 10% Iberian mean that 10% of your ancestry came from actual Iberian Neolithic migrants?

All the same components as in Mesolithic Europeans were found, but also:

- North_Caucasian
- South_Central_Asian
- Volga-Ural

Plus some traces of other admixtures that do not consistently show up, like Amerindian and African.

Also, I was wondering. How come the Steppe ancestry is ending up in the Mesolithic European components? Is there a close genetic link between the Yamna Culture and the European HG?

 

[Northener]

Thanks Maciamo it makes K36 more clear. But when I counted my components I got doubts:

I. HG 69,4%
II. EEF 19%
III Corded Ware/ Bronze Age 7,4%

Is this possible?

...................
My calculation:
Mesolithic hunter-gatherer components
- East_Central_Euro 8
- Eastern_Euro 3,6
- Fennoscandian 13,6
- North_Atlantic 18,4
- North_Sea 25

I. 68,6% HG + 0,8% Basque= 69,4%
These were found in WHG (Spain), SHG (Sweden) and EHG (Yamna, Afanasevo, Corded Ware).

Neolithic farmer components
- Arabian
- Armenian
- East_Balkan 6,2
- East_Med
- Iberian 6,4
- Italian 6,4
- Near_Eastern
- North_African
- Omotic
- West_Med

II. 19% EEF

The "French" component III. Corded Ware Denmark (11.5%), and even higher in Corded Ware Poland (26%) and in Nordic Bronze Age Denmark (31.5%), but then sharply decreases in Iron Age Sweden (6%). 7,4%

The Basque component was found in WHG (8% and 11%), in Funnelbeaker Sweden (2% and 28%), in Chalcolithic Hungary (8%), in Corded Ware Poland (14.5%), Sintashta (1.5%), and in the Nordic Bronze Age (3%). 0,8%

 

[clarbg]

i got 1.47% south asian on k36 and noticed most people of my ethnicity either scored 0 or less than 1%. does this mean it's more accurate for indian ancestry? i supposedly may have distant gypsy ancestry. on most of the calculators i get 1-4%.

 

[Capitalis]

I thank Maciamo for his work, which I found very interesting.

I added up the components for all areas of Europe on Lukasz's previous K36 update and share a sample of the results here, which generally support the study summary quoted below.


Mesolithic-Neolithic-Bronze Age

N/E Finland: 95-0-5
Lithuania: 87-5-8
...
Scotland: 60-22-18
N Holland: 67-16-17
...
C Greece: 15-68-17
S Italy: 3-85-12

Me (Eng/Ire/Wal): 63-18-19


From: Wikipeda > Neolithic Europe > Genetic Studies

Dupanloup performed an admixture analysis based on several autosomal loci, mtDNA and NRY haplogroup frequencies. The study was based on the assumption that Basques were modern representatives of Palaeolithic hunter-gatherers’ gene pool, and Near Eastern peoples were a proxy population for Neolithic farmers. Subsequently, they used admixture analysis to estimate the likely components of the contemporary European gene pool contributed by the two parental populations whose members hybridized at a certain moment in the past. The study suggested that the greatest Near Eastern admixture occurs in the Balkans (~80%) and Southern Italy (~60%), whilst it is least in peoples of the British Isles (estimating only a 20% contribution). The authors concluded that the Neolithic shift to agriculture entailed major population dispersal from the Near East.^[16]

 

[Angela]

I thank Maciamo for his work, which I found very interesting.

I added up the components for all areas of Europe on Lukasz's previous K36 update and share a sample of the results here, which generally support the study summary quoted below.


Mesolithic-Neolithic-Bronze Age

N/E Finland: 95-0-5
Lithuania: 87-5-8
...
Scotland: 60-22-18
N Holland: 67-16-17
...
C Greece: 15-68-17
S Italy: 3-85-12

Me (Eng/Ire/Wal): 63-18-19


From: Wikipeda > Neolithic Europe > Genetic Studies

See:

See also:
https://www.eupedia.com/forum/threads/34850-Important-papers-for-newbies-to-Population-Genetics

 

[Capitalis]

Thanks. So my numbers are way off then, I guess?

Edit: Or, most of what I counted as Mesolithic came in during the Bronze Age, as per Maciamo's note that the components were the same.

 

[Angela]

Thanks. So my numbers are way off then, I guess?

Edit: Or, most of what I counted as Mesolithic came in during the Bronze Age, as per Maciamo's note that the components were the same.

For a rough estimate, Capitalis, I usually apply a 60% EHG/SHG factor to Yamnaya on the Haak chart,with the remaining 40% being "Caucasus" like or Iran Chl like.

 

[Capitalis]

Thanks again. I'll read my way through your newbies thread, looks great.

 

[Dibran]

Here is the frequency of a few components by prehistoric culture.


Central_Euro (central-north European)

- Corded Ware Poland: 25%
- Corded Ware Denmark: 16%
- Sintashta: 12%
- Corded Ware Germany (Esperstedt): 9%
- Iron Age England : 7.5%
- Iron Age Sweden : 7.5%
- Mesolithic Sweden: 6%
- Yamna: 1%
- Nordic Bronze Age Denmark: 0%
- Afanasievo : 0%
- Mesolithic Hungary: 0%
- Mesolithic Karelia: 0%
- Mesolithic Luxembourg: 0%
- Mesolithic Samara: 0%
- Mesolithic Spain: 0%
- Neolithic Europe: 0%
- Palaeolithic Siberia: 0%

East_Central_Euro (very Mesolithic European, perhaps linked to mtDNA U4 or U5)

- Mesolithic Sweden: 43.5%
- Mesolithic Hungary: 30%
- Mesolithic Samara: 26%
- Corded Ware Poland: 23%
- Mesolithic Spain: 20.5%
- Mesolithic Luxembourg: 20.5%
- Afanasievo : 16.5%
- Sintashta: 15%
- Yamna: 10%
- Mesolithic Karelia: 9%
- Iron Age Sweden : 7.5%
- Nordic Bronze Age Denmark: 6%
- Corded Ware Denmark: 5.5%
- Palaeolithic Siberia: 4.5%
- Iron Age England : 3.5%
- Corded Ware Germany (Esperstedt): 3%
- Neolithic Europe: 0%

Eastern_Euro (Ancient North Asian?)

- Mesolithic Karelia: 23.5%
- Palaeolithic Siberia: 17.5%
- Mesolithic Samara: 17%
- Yamna: 15%
- Corded Ware Germany (Esperstedt): 12.5%
- Sintashta: 12%
- Mesolithic Spain: 11%
- Mesolithic Luxembourg: 10%
- Iron Age Sweden : 7.5%
- Iron Age England : 5%
- Mesolithic Hungary: 3%
- Mesolithic Sweden: 2.5%
- Afanasievo : 0%
- Nordic Bronze Age Denmark: 0%
- Corded Ware Poland: 0%
- Corded Ware Denmark: 0%
- Neolithic Europe: 0%

Fennoscandian (Mesolithic Europe and North Asia, perhaps linked to mt-haplogroup U)

- Mesolithic Luxembourg: 39%
- Mesolithic Samara: 36%
- Mesolithic Spain: 33.5%
- Mesolithic Karelia: 33%
- Mesolithic Hungary: 32.5%
- Palaeolithic Siberia: 21.5%
- Nordic Bronze Age Denmark: 19.5%
- Corded Ware Germany (Esperstedt): 18%
- Iron Age Sweden : 15%
- Afanasievo : 14.5%
- Yamna: 14.5%
- Mesolithic Sweden: 10%
- Iron Age England : 9.5%
- Corded Ware Denmark: 8%
- Sintashta: 7.5%
- Corded Ware Poland: 0%
- Neolithic Europe: 0%

French (especially Baltic; similar to Central_Euro)

- Nordic Bronze Age Denmark: 31.5%
- Corded Ware Poland: 26%
- Corded Ware Denmark: 11.5%
- Sintashta: 10%
- Iron Age England : 9%
- Iron Age Sweden : 6.5%
- Afanasievo : 2.5%
- Corded Ware Germany (Esperstedt): 2%
- Yamna: 0%
- Mesolithic Hungary: 0%
- Mesolithic Karelia: 0%
- Mesolithic Luxembourg: 0%
- Mesolithic Samara: 0%
- Mesolithic Spain: 0%
- Mesolithic Sweden: 0%
- Neolithic Europe: 0%
- Palaeolithic Siberia: 0%

North_Atlantic (mostly northern European)

- Sintashta: 21.5%
- Iron Age Sweden : 19.5%
- Mesolithic Sweden: 19.5%
- Iron Age England : 19%
- Afanasievo : 16.5%
- Corded Ware Denmark: 15.5%
- Corded Ware Germany (Esperstedt): 15%
- Mesolithic Hungary: 10%
- Yamna: 9%
- Mesolithic Luxembourg: 9%
- Mesolithic Spain: 8%
- Mesolithic Karelia: 7%
- Mesolithic Samara: 0%
- Nordic Bronze Age Denmark: 0%
- Corded Ware Poland: 0%
- Neolithic Europe: 0%
- Palaeolithic Siberia: 0%

North_Sea (Mesolithic but especially SHG)

- Corded Ware Denmark: 28%
- Afanasievo : 26%
- Iron Age Sweden : 24.5%
- Corded Ware Germany (Esperstedt): 20%
- Mesolithic Hungary: 19.5%
- Mesolithic Sweden: 18.5%
- Iron Age England : 16.5%
- Mesolithic Spain: 16%
- Sintashta: 15%
- Mesolithic Luxembourg: 13%
- Mesolithic Samara: 11.5%
- Nordic Bronze Age Denmark: 11.5%
- Yamna: 11.5%
- Mesolithic Karelia: 5%
- Palaeolithic Siberia: 0.5%
- Corded Ware Poland: 0 %
- Neolithic Europe: 0%

Do you plan on doing frequency of components by prehistoric culture for the rest of the components?

 

[Capitalis]

I read my way through the linked Newbie papers but I still have questions and no fear of looking stupid, so:

Although the North Sea/Atlantic components were present in NW Europe in the Mesolithic (see Maciamo's post above), the message that I got from some of the papers and yourself Angela is that this majority part of my K36 ancestry has been inherited from a population that spread from the steppes to NW Europe, replacing up to 95% of the Neolithic British population according to one paper. So were these steppe-descended people carrying the North Sea/Atlantic components?

A logical consequence of population replacement is that I can't have inherited these components from pre-Bronze Age NW European inhabitants. There isn't enough WHG on that chart to account for the ~45-50% of K36 ancestry that NW Europeans derive from the North Sea/Atlantic components, so the Yamnaya part must contain the rest, I assume, as the components don't appear to be present in the Neolithic samples. I'm confused because these components decline rapidly from East Germany out to the Baltics and Ukraine, yet they too derive a similar amount of their ancestry from Yamnaya.

I realise K36 isn't everyone's cup of tea but I'm mildly invested in it for the time being, so please humour me.

I'll post one question at a time. Thanks for any replies.

 

[Duarte]

Thanks for the help Maciamo. I had a hard time understanding my results from GEDmach Eurogenes K36. Now everything has become clearer to me after following the rating you suggested. In fact, with the adoption of its classification my results gain logic whem I compare with the other GEDmatch calculators and the companies that operate in the autosomal test market.
Below my results, acording your suggested classification:

1) EUROPEAN COMPONENTS: 94.48

Mesolithic hunter-gatherer components: 16.83

- East_Central_Euro: 0.15
- Eastern_Euro: 2.25
- Fennoscandian: 1.18
- North_Atlantic: 7.79
- North_Sea: 5.46

Neolithic farmer components: 64.79

- Arabian: 0.00
- Armenian: 0.79
- East_Balkan: 2.97
- East_Med: 0.00
- Iberian: 21.51
- Italian: 25.19
- Near_Eastern: 1.19
- North_African: 5.34
- Omotic: 1.25
- West_Med: 6.55

French: 7.86

Basque: 5.00

Bronze Age Steppe components: 0.00

- North_Caucasian: 0.00
- South_Central_Asian: 0.00
- Volga-Ural: 0.00

2) SUB-SAHARAN COMPONENTS: 4.38
- East_African: 0.76
- Pygmy: 0.65
- West_African: 2.97

3) AMERINDIAN/OCEANIAN COMPONENTS: 1.12
- Amerindian: 0.76
- Oceanian: 0.36

 

[tuț]

I'm curious what 1,16% Indo-Chinese can mean for K36? I find it kind of weird... Vietnamese?

 

[Joey37]

Hunter-Gatherer: 54.44
-East Central Euro 10.22
-East Euro 3.22
-Fennoscandian 11.01
-North Atlantic 13.71
-North Sea 16.28
Neolithic Farmer: 30.29
-East Balkan 2.22
-Iberian 10.64
-Italian 12.56
-West Med 4.87
French: 5.66
Basque: 2.52
Bronze Age Steppe components: 2.75
-North Caucasian 1.67
-Volga-Ural 1.08
I am predominantly of northwest European descent.

 

[tuț]

Italian ..............................13,65% f
Iberian .............................9,89% f
Central_Euro .................... 9,63%
East_Balkan ..................... 9,45% f
East_Central_Euro .............8,98% h
Eastern_Euro ....................7,92% h
North_Sea ........................7,66% h
East_Med .........................6,36% f
West_Caucasian ................5,56%
Near_Eastern ....................4,56% f
West_Med ........................3,65% f
French .............................3,59%
Basque ........................... 3,38%
Arabian ............................2,9% f
Indo-Chinese .................1,16%
North_African ...................0,71% f
Armenian ........................ 0,45% f
Amerindian ......................0,21%
South_Central_Asian .........0,12% s
North_Atlantic ..................0,09% h
Fennoscandian ..................0,04% h
Volga-Ural ........................0,03% s
Malayan ...........................0,01%

Farmer ...... 51,62%
Hunter ...... 24,69%
Stepe ......... 0,15%
The rest .... 23,54%

I saw another ~20 results from the center and north of the Balkans and the results are very similar to this one. What is exotic is only Indo-Chinese 1,16% that is not supported by other percentages in South - East Asia (South-Asian, South-Chinese, Oceanian, except Malayan but only with 0.01%).
It is excluded that at least in the last 7-8 generations there have been ancestors from other parts than from Romania.
Do you think that these 1.16% Indo-Chinese can be a real heritage or are somehow missed in the result being a mistake?

 

[Variel]

European: 100%
Gatherer: 61,29% (North Sea 25.09, Atlantic 18.48, Fennoscan 13.44, ECE 3.78, EE 0,50)
Farmer: 18,66% (Italian 13.74, East Balkan 4.25 Iberian 0,.65, West Med 0.02)
Steppe: 0,99% (Volga 0.99)
Other: 19,04% (French 6.79, Basque 3.89, CE 8.36)