PDA

View Full Version : Help with Eurogenes K13 Oracle results - French ancestry?



Oceane81
09-16-2018, 11:57 PM
Hi,

I'm wondering if someone could help me interpret my father's K13 results. As far as I know, our ancestry is mainly French (various northern and southern regions). However, it seems like his results are shifted more to the south than I would have expected. I assumed our southern lines were more distant. I also find it strange that he has Finnish as a secondary population. The Scandinavian regions make sense, since we do have lines in northern France. I'm wondering if the Finnish could be a proxy for something else? I compared his components with the French reference population and he does have slightly more East Med (15.90). However, what I find confusing is the low West Asian (0.33). I have 4.07, which is a bit lower than the French sample, but in line with the average, I think.

If anyone could help me make sense of this I would appreciate it. Thank you.

Admix Results (sorted):

# Population Percent
1 North_Atlantic 41.7
2 Baltic 20.42
3 West_Med 18.37
4 East_Med 15.9
5 East_Asian 0.94
6 Amerindian 0.71
7 Red_Sea 0.61
8 Siberian 0.47
9 Oceanian 0.38
10 West_Asian 0.33
11 Northeast_African 0.1

Mixed Mode Population Sharing:

Primary Population (source) Secondary Population (source) Distance
1 70.4% Spanish_Valencia + 29.6% Southwest_Finnish 5.82
2 74.4% Spanish_Valencia + 25.6% La_Brana-1 5.83
3 65.7% Spanish_Valencia + 34.3% North_Swedish 5.83
4 70.1% Spanish_Murcia + 29.9% Southwest_Finnish 6
5 74.1% Spanish_Valencia + 25.9% Finnish 6.03
6 65.5% Spanish_Murcia + 34.5% North_Swedish 6.13
7 74.9% Spanish_Cataluna + 25.1% Southwest_Finnish 6.16
8 78.6% Spanish_Cataluna + 21.4% La_Brana-1 6.19
9 78.2% Spanish_Cataluna + 21.8% Finnish 6.25
10 73.9% Spanish_Murcia + 26.1% Finnish 6.3
11 74.4% Spanish_Murcia + 25.6% La_Brana-1 6.32
12 61.2% Southeast_English + 38.8% Tuscan 6.32
13 53.4% Southeast_English + 46.6% North_Italian 6.34
14 78.1% Spanish_Cataluna + 21.9% Belorussian 6.35
15 74.7% Spanish_Murcia + 25.3% Estonian 6.36
16 75.1% Spanish_Valencia + 24.9% Estonian 6.36
17 75.2% Southeast_English + 24.8% Algerian_Jewish 6.37
18 68.9% Portuguese + 31.1% North_Swedish 6.39
19 71% Spanish_Cataluna + 29% North_Swedish 6.40
20 74.7% Southeast_English + 25.3% Italian_Jewish 6.42



Using 2 populations approximation:
1 50% North_Italian +50% Southeast_English 7.356437


Using 3 populations approximation:
1 50% Portuguese +25% Southwest_Finnish +25% Spanish_Valencia 7.093826


Using 4 populations approximation:

1 North_Swedish + Southeast_English + Spanish_Valencia + Tuscan 6.427474
2 La_Brana-1 + Spanish_Valencia + Spanish_Valencia + Spanish_Valencia 6.456249
3 La_Brana-1 + Spanish_Murcia + Spanish_Valencia + Spanish_Valencia 6.554781
4 Hungarian + Southeast_English + Spanish_Valencia + Spanish_Valencia 6.557671
5 Ashkenazi + North_Swedish + Southeast_English + Spanish_Aragon 6.568531
6 La_Brana-1 + Spanish_Cataluna + Spanish_Valencia + Spanish_Valencia 6.576902
7 Hungarian + Southeast_English + Spanish_Murcia + Spanish_Valencia 6.621439
8 North_Swedish + Southeast_English + Spanish_Murcia + Tuscan 6.647042
9 La_Brana-1 + Portuguese + Spanish_Valencia + Spanish_Valencia 6.649552
10 French_Basque + Italian_Jewish + Southeast_English + Southwest_Finnish 6.684114
11 Danish + Hungarian + Spanish_Valencia + Spanish_Valencia 6.691012
12 La_Brana-1 + Spanish_Cataluna + Spanish_Murcia + Spanish_Valencia 6.696943
13 North_Swedish + Southeast_English + Spanish_Valencia + West_Sicilian 6.700554
14 Ashkenazi + Norwegian + Southeast_English + Spanish_Valencia 6.702642
15 La_Brana-1 + Spanish_Castilla_Y_Leon + Spanish_Valencia + Spanish_Valencia 6.716030
16 French + Southwest_Finnish + Spanish_Valencia + Spanish_Valencia 6.716554
17 Hungarian + Portuguese + Southeast_English + Spanish_Valencia 6.721418
18 La_Brana-1 + Spanish_Murcia + Spanish_Murcia + Spanish_Valencia 6.736481
19 Ashkenazi + Southeast_English + Spanish_Aragon + Swedish 6.737554
20 Ashkenazi + Southeast_English + Southeast_English + Southwest_French 6.738945

msmarjoribanks
09-17-2018, 07:01 AM
Yeah, the populations in Gedmatch always can be proxies for other things. I wouldn't assume Finnish means anything other than an explanation for your Baltic numbers and not a very good one. It's just computer generated.

What single populations did you get and how close?

Best way to work with Eurogenes K13 is to compare to the spreadsheet (and it only has two French categories, French and French Basque, which is limited).

French spreadsheet (compared to your results, second) says:

N Atl: 42.6 vs. 41.7 -- close
Baltic: 17.5 vs. 20.4 -- yours is high
West Med: 19.5 vs. 18.4 -- yours is close, but a bit low
West Asian: 5 vs. .33 -- v. low
East Med: 10.4 vs. 15.9 -- high, but likely due to low W Asian, the test can confuse these

It doesn't look like a southern shift to me, you are simply getting Spain as a base vs. France (sometimes the computer on these picks more extreme populations) and then a compensation with northern inputs like Finnish. It looks like you have a bit less W Med and more Baltic than the average French result, which could be some Northern French or even Nordic input from far back. The W Asian/E Med should probably be combined so I wouldn't make much of that.

Oceane81
09-17-2018, 10:47 AM
Hi,

Thank you for your response. I'm not sure, since I've seen northern French results and they don't look like this at all. They usually have something Scandinavoan or English as a base (danish, Swedish, southeast English) in mixed mode. There's some admixture there, but overall there's more West Med influence even in the 4 populations. This isn't typical of a northern result. His top single population is French, but the distance is quite high at 7. In fact, all of the distances are high, much more than mine, which I find interesting.

Although he's closer to West German and South Dutch, his genetic distance to English is low, and he has more Mediterranean regions in his single populations than I do. So I don't think it's northern. If it was, his distance to south dutch would be much closer. There seems to be a lot of admixture, but I can't determine if it's southeast or southwest.


Single Population Sharing:

# Population (source) Distance
1 French 7.28
2 West_German 9.38
3 South_Dutch 9.71
4 Spanish_Cataluna 10.63
5 Portuguese 11.34
6 Spanish_Castilla_Y_Leon 11.75
7 Spanish_Galicia 12.02
8 Spanish_Valencia 12.38
9 Spanish_Murcia 12.6
10 Austrian 13.34
11 Southeast_English 13.44
12 Spanish_Extremadura 13.59
13 Spanish_Cantabria 13.78
14 Southwest_French 14.04
15 Spanish_Castilla_La_Mancha 14.17
16 North_Italian 14.97
17 East_German 15.11
18 Spanish_Andalucia 15.23
19 Southwest_English 15.34
20 Spanish_Aragon 15.5

palamede
09-17-2018, 11:03 AM
I will be cautious with the oracle giving several composants due to the mode of calculation which doesn't group by association of SNPs because too costly to find and to calculate. They use a simpler calculation mode to obtain a set of SNPS the nearest possible of your result. It is only useful if at least one of the components is in your ancestry, it is an anchor, then you can hope the other components make sense in your ancestry .

The french samples of the dataset are dominated by South-East France and a lot less by West and North French.Therefore more East Med and Red Sea compared to the average of French people. you have an excess of East Med compared to a lot of French people, except those of the Mediterranean coast and Corsica

I am from North of France, but compared to France people, I have a little excess for West Asia, other calculators give me an excess for Caucasus (also part of Eurogenes K13 and K15 West Asian), I haven't found ancestry which explains this excess for K13 and K15)

Here is my K13 (FTDNA family finder dataset) : North Atlantic 43,92 Baltic 20,17 West Med 18,68 West Asian 8,57 East Med 4.83 Red Sea 1,09 SE Asian 0,92,Amerindian 1,20 Oceanian 0,62 ;
For European populations, I prefer K15 :
North Sea 26,27 Atlantic 28,3 Baltic 10,14, Eastern Europe 7,85 West Med 15,00 West Asian 8.39 East Med 1,51 Red Sea 0,63 SE Asian 0,37 Amerindian 1,12 Oceanian 032

You can access to the French forum and look for help https://anthrogenica.com/forumdisplay.php?162-French
If you are not sure to write an understandable french , no problem to write in english.

In the oracle 4 composants, the differences between the 20 combinations are wery weak, one of the more interesting is
20 Ashkenazi + Southeast_English + Southeast_English + Southwest_French 6.738945
Southeast_English is not far from North-West France. In the French forum several of the North West France participants are often given as South English.

Oceane81
09-17-2018, 11:22 AM
Thank you. So it seems that the biggest difference is the higher East Med. It's still high in the K15 (13.48). Baltic is lower in that one compared to the K13. Could it be northern Italian admixture?

Admix Results (sorted):

# Population Percent
1 North_Sea 34.77
2 Atlantic 17.21
3 West_Med 16.83
4 East_Med 13.48
5 Baltic 8.39
6 Eastern_Euro 7.32
7 Southeast_Asian 0.76
8 Northeast_African 0.51
9 Amerindian 0.51
10 Siberian 0.12
11 Oceanian 0.1

Ruderico
09-17-2018, 11:42 AM
Didn't people from NE France exhibit some increased East Med in K15 when compared to other French? I might be wrong, but I remember it being reported around here

Oceane81
09-17-2018, 11:50 AM
Southeast_English is not far from North-West France. In the French forum several of the North West France participants are often given as South English.

Yes, but those who are closer to northwest France get that as the top population in mixed mode. It seems like a mix between north and south to me. I'm not sure if it's just a problem with the SNPs. I don't get that and my mother (who has more Norman and northeast ancestry) is shifted more towards the English regions in her oracles.

Thanks for your help.

msmarjoribanks
09-17-2018, 12:24 PM
I'm still skeptical about the East Med being meaningful when the West Asian is low in just about the same proportion. True heightened E Med would come with heightened W Asian -- here it looks like W Asian could be classified as E Med.

Higher Baltic/lower W Med is the only real difference. I think that's an eastern or northern tilt -- normally that would come with heightened N Alt too, but not here, so maybe German, Scandinavian, not British Isles.

Could be a local result within France -- there's got to be more variation than the spreadsheet allows for given the limited French samples.

palamede
09-17-2018, 12:44 PM
duplicate

palamede
09-17-2018, 12:52 PM
[QUOTE=Oceane81;490628]Thank you. So it seems that the biggest difference is the higher East Med. It's still high in the K15 (13.48). Baltic is lower in that one compared to the K13. Could it be northern Italian admixture?

..................

Northern Italian or Switzerland is possible .

We have no equivalence between Baltic K13 and Baltic 15. Sometimes, we must defy about the labels, they are given by the creator of the calculator to define the component at the end of the process and it is not good always. In K13, we have 3 "European" components North-Atlantic, Baltic and West Med; and in K15 we have 5 "European" components NorthSea, Atlantic, Baltic Eastern Europe and West Med, and that leads to some changes in non-European components between K13 and K15 for instance K15 West Asian is more Caucasian than K13 West Asian.

I guess Ruderico is right : East Med is higher in East France than in North and West France, I have no result for East France and Switzerland, but for West Germany
1) North Sea 34.00 2) Atlantic 21.83 3) West Med 11,13 4) Baltic 10.46 5) East Europe 8.06 6) East Med 6.33 7) West Asian 5.00 8) Red Sea 1.02 9) South Asian 0.72 ....

I say again you should ignore the admixture without at least one component which corresponds to your known ancestry.

I am half a Norman, half a Walloon. the middle way is Vermandois-Picardy-North France.

Oceane81
09-17-2018, 04:56 PM
I
Higher Baltic/lower W Med is the only real difference..

The difference is not very big for the Baltic, from 17 to 20. As well as the West Med, 18.40 to 19.50. I looked at the Basque French sample and they have 0.35 West Asian, which is interesting and almost like his 0.33. It could be some sort of extreme admixture from the East. The low Atlantic is also strange. I agree, it's definitely not British Isles. In other calculators his Celtic is quite low. I still think this isn't a normal result for someone with mostly ancestry from northern and central France, but I could be wrong.

Thanks for your help!

Oceane81
09-17-2018, 05:01 PM
[QUOTE=Oceane81;490628]

I guess Ruderico is right : East Med is higher in East France than in North and West France, I have no result for East France and Switzerland, but for West Germany
1) North Sea 34.00 2) Atlantic 21.83 3) West Med 11,13 4) Baltic 10.46 5) East Europe 8.06 6) East Med 6.33 7) West Asian 5.00 8) Red Sea 1.02 9) South Asian 0.72 ....

That result looks very similar, but in his K15 he has higher West Med and East Med. The Baltic is also slightly lower. It's probably shifted towards Eastern France. Very interesting.

Population Percent
1 North_Sea 34.77
2 Atlantic 17.21
3 West_Med 16.83
4 East_Med 13.48
5 Baltic 8.39
6 Eastern_Euro 7.32
7 Southeast_Asian 0.76
8 Northeast_African 0.51
9 Amerindian 0.51
10 Siberian 0.12
11 Oceanian 0.1

Oceane81
09-17-2018, 05:30 PM
My K13 and K15 for comparison:

Admix Results (sorted):

# Population Percent
1 North_Atlantic 42.92
2 Baltic 20.61
3 West_Med 19.07
4 East_Med 10.34
5 West_Asian 4.07
6 East_Asian 1.46
7 Amerindian 1.33
8 Siberian 0.1

K15

Admix Results (sorted):

# Population Percent
1 North_Sea 32.88
2 Atlantic 22.16
3 West_Med 15.81
4 Eastern_Euro 9.9
5 East_Med 8.75
6 Baltic 6.83
7 West_Asian 1.62
8 Southeast_Asian 1.06
9 Amerindian 0.9

Helves
09-17-2018, 07:09 PM
What kind of raw data are you using? 23andme v5?

Oceane81
09-17-2018, 07:31 PM
Yes, I'm wondering if that might not be the problem... It seems to shift things more east than what would be expected. I'm testing with FTDNA, so I'm looking forward to seeing if there's a difference.

AJL
09-20-2018, 08:47 PM
Moved to Open-Source Projects.