Who is more Germanic between Swiss Germans and Scots?

[Septentrion]

I am talking genetically and phenotypically, I am NOT taking into account culture because then it would be an easy win for Swiss Germans.

I would say the Scottish particularly the Lowland Scottish who have Anglo - Saxon blood as over half of them have Germanic paternal lineages. Blue eyes and a very fair skin type are much more common in Scottish population than in the Swiss one. Highlander Scottish although are some kind «Celts of Celts» pure leading all the way to the Beaker Folk. Swiss Germans are more Mongrel with Celtic ( Helvetic), Roman blood admixtures. This is why they are also genetically further away from the Germanic Core ( Scandinavia ) than Scots.

[Grace O'Malley]

It's closer to 42%, with more indepth "overfitted" models, but still more than Scots, who are around 34-37%, on par with England. 20% is far too low for Scotland, that's around Ireland's level.

Scotland is obviously more NW Euro/Bell Beaker overall though.

Not saying I disagree with you but why do Irish cluster so closely with Scots and English? Is it the Beaker ancestry that balances it or did Germanic input not change their overall genetics. Anyway it might be an impossible question to answer?

[Grace O'Malley]

You don´t like Czechia and you don´t like the Bronze Age? Because you changed also this...

But ok, would you use these references in a model to compare the Germanic input in Scottish and Swiss?

DEU_MA,0.1223596,0.1303939,0.061169,0.048773,0.039792,0.0199408,0.010975,0.0052151,0.0013295,-0.0024966,-0.003735,0.001109,-0.0091576,-0.0038398,0.0161643,-0.0008352,-0.0133511,0.0032684,0.0041354,0.0040271,0.0060019,0.0037342,-0.0007273,0.011146,-0.0004429
Switzerland_IA,0.125205,0.153345,0.050534,-0.003876,0.049548,-0.005578,0.0047,-0.002077,0.014726,0.033896,0.001786,0.016335,-0.016353,-0.008945,0.001764,-0.001591,0.009518,0.001647,0.002514,-0.006128,0.005241,0.005688,-0.012078,-0.004338,0.000958
ITA_Rome_Imperial,0.1039821,0.1495156,-0.0235307,-0.0574065,0.0045265,-0.0204055,-0.0011946,-0.0051488,0.0006604,0.0196549,0.0034575,0.0025539,-0.0040602,-0.0014737,-0.0081715,-0.0014474,0.0035992,0.000454,0.0012178,-0.0032854,-0.0025579,0.0020454,-0.0006985,-0.0004845,0.0004141
ISL_Viking_Age_Pre_Christian,0.1281142,0.1294236,0.0681749,0.0618366,0.0366562,0.0195534,0.0069459,0.0088458,0.0002728,-0.0021463,-0.0048718,0.009608,-0.0101914,-0.0155362,0.0242939,0.0120656,0.0019703,0.0065033,0.002947,0.0051831,0.0056566,0.004053,0.0024102,0.0136563,0.0015302
Scotland_LBA,0.1297585,0.133542,0.0630732,0.0527298,0.0351605,0.0193132,0.0061102,0.0030575,0.001892,-0.0027338,-0.0060082,0.0064818,-0.0180252,-0.0164115,0.024226,0.0058672,-0.0143422,0.0036108,0.0032995,0.0005313,0.0056462,0.0044822,-0.0003393,0.0042172,0.0041315

Or would you make 2 different models to see the Germanic input in every group, changing the BA/IA local references?:

For Swiss people:

DEU_MA,0.1223596,0.1303939,0.061169,0.048773,0.039792,0.0199408,0.010975,0.0052151,0.0013295,-0.0024966,-0.003735,0.001109,-0.0091576,-0.0038398,0.0161643,-0.0008352,-0.0133511,0.0032684,0.0041354,0.0040271,0.0060019,0.0037342,-0.0007273,0.011146,-0.0004429
Switzerland_IA,0.1265966,0.1223149,0.0592498,0.0698398,0.024996,0.0205758,0.0038906,0.0014357,-0.0044768,-0.0130401,-0.0025803,0.0016151,-0.0079617,-0.0124013,0.0223186,0.0107692,-0.0020861,0.0008588,0.0025558,0.0044882,0.0072788,0.0010992,0.0006572,0.0054892,-0.0010112
ITA_Rome_Imperial,0.1039821,0.1495156,-0.0235307,-0.0574065,0.0045265,-0.0204055,-0.0011946,-0.0051488,0.0006604,0.0196549,0.0034575,0.0025539,-0.0040602,-0.0014737,-0.0081715,-0.0014474,0.0035992,0.000454,0.0012178,-0.0032854,-0.0025579,0.0020454,-0.0006985,-0.0004845,0.0004141

For Scottish people:

DEU_MA,0.1223596,0.1303939,0.061169,0.048773,0.039792,0.0199408,0.010975,0.0052151,0.0013295,-0.0024966,-0.003735,0.001109,-0.0091576,-0.0038398,0.0161643,-0.0008352,-0.0133511,0.0032684,0.0041354,0.0040271,0.0060019,0.0037342,-0.0007273,0.011146,-0.0004429
Scotland_LBA,0.1297585,0.133542,0.0630732,0.0527298,0.0351605,0.0193132,0.0061102,0.0030575,0.001892,-0.0027338,-0.0060082,0.0064818,-0.0180252,-0.0164115,0.024226,0.0058672,-0.0143422,0.0036108,0.0032995,0.0005313,0.0056462,0.0044822,-0.0003393,0.0042172,0.0041315
ITA_Rome_Imperial,0.1039821,0.1495156,-0.0235307,-0.0574065,0.0045265,-0.0204055,-0.0011946,-0.0051488,0.0006604,0.0196549,0.0034575,0.0025539,-0.0040602,-0.0014737,-0.0081715,-0.0014474,0.0035992,0.000454,0.0012178,-0.0032854,-0.0025579,0.0020454,-0.0006985,-0.0004845,0.0004141

And what must we do with this other reference?:

ISL_Viking_Age_Pre_Christian,0.1281142,0.1294236,0.0681749,0.0618366,0.0366562,0.0195534,0.0069459,0.0088458,0.0002728,-0.0021463,-0.0048718,0.009608,-0.0101914,-0.0155362,0.0242939,0.0120656,0.0019703,0.0065033,0.002947,0.0051831,0.0056566,0.004053,0.0024102,0.0136563,0.0015302

I would say that it is unnecessary, since it would be redundant, because we already have DEU_MA to measure the Germanic percentage.

It could even be a better reference to measure the germanic contribution than CHE_IA...but well, it´s your model and are your references.

This is what I get with those models.

First model.


Target: Grace_scaled
Distance: 0.0216% / 0.02163678
54.9 Scotland_LBA
33.2 ISL_Viking_Age_Pre_Christian
8.6 DEU_MA
3.3 Switzerland_IA

With just DEU and Scotland_LBA


Target: Grace_scaled
Distance: 0.0232% / 0.02315752
81.0 Scotland_LBA
19.0 DEU_MA

With the ISL_Viking


Target: Grace_scaled
Distance: 0.0218% / 0.02179288
56.8 Scotland_LBA
32.2 ISL_Viking_Age_Pre_Christian
10.3 DEU_MA
0.7 ITA_Rome_Imperial

With DEU removed.

Target: Grace_scaled
Distance: 0.0219% / 0.02191259
63.4 Scotland_LBA
35.3 ISL_Viking_Age_Pre_Christian
1.3 ITA_Rome_Imperial

So why do I get such a high amount of Germanic? I'm not sure whether this will ever be solved unless there are more studies. I think different areas even in places like Ireland are not all uniform.

[Defcon2]

So why do I get such a high amount of Germanic? I'm not sure whether this will ever be solved unless there are more studies. I think different areas even in places like Ireland are not all uniform.

Have you tried using R/Nmonte on unscaled models to try to better discriminate the components?

[princeton90]

Swiss Germans

[Grace O'Malley]

Have you tried using R/Nmonte on unscaled models to try to better discriminate the components?

No I've never used R. With Unscaled this is what I get.

Target: Grace_unscaled
Distance: 1.3970% / 0.01397028
55.2 Scotland_LBA
44.8 ISL_Viking_Age_Pre_Christian

Target: Grace_unscaled

Distance: 1.3956% / 0.01395626
54.4 Scotland_LBA
44.6 ISL_Viking_Age_Pre_Christian
1.0 ITA_Rome_Imperial

[Defcon2]

No I've never used R. With Unscaled this is what I get.

Target: Grace_unscaled
Distance: 1.3970% / 0.01397028
55.2 Scotland_LBA
44.8 ISL_Viking_Age_Pre_Christian

Target: Grace_unscaled

Distance: 1.3956% / 0.01395626
54.4 Scotland_LBA
44.6 ISL_Viking_Age_Pre_Christian
1.0 ITA_Rome_Imperial

The reason I tell you R/Nmonte is because of its penalty, unscaled models need penalty. Supposedly with ADD DIST COL: 0.25x I would simulate this, but it doesn't work.

[gixajo]

Where? When will you stop acting like a passive-agressive bitch?

I will better no comment this.

The model with five references will produce overfitting, and it indeed does:

Target: Swiss_German
Distance: 1.1769% / 0.01176925
36.0 DEU_MA
33.4 Switzerland_IA
12.8 Scotland_LBA
10.6 ITA_Rome_Imperial
7.2 ISL_Viking_Age_Pre_Christian

That happens because you have three extremely closely related populations in the same model, as i've been saying repeatedly for decades. An overfitted model should be sent to thrash.

Agree that´s because I said that would be redundant.

You always should use more proximal populations. Does it makes sense to model Scots with Medieval Bavarians?


There isn't.

Depending the purpose of the model, and depending more things, but that is anotehr discussion.

Thank you very much for showing how you work.

And to summarize, could you post the sources of the model or models (models with all the sources, not just results, and please tell us where from you get the references if you change any label again) that you would use to compare the Germanic contribution of Scots and Swiss? (To answer the question of the thread, or at least, give your version of the matter through the results of those models).

[gixajo]

No I've never used R. With Unscaled this is what I get.

l

The reason I tell you R/Nmonte is because of its penalty, unscaled models need penalty. Supposedly with ADD DIST COL: 0.25x I would simulate this, but it doesn't work.

You can use also this one: https://yk.github.io/ancestry/

It´s more easy to use and more intuitive than R directly, and you have more options to configurate penalties (or penalization, I don´t know well which word use in English).

When I use it, I use "Adam medium" option, but you can play with different configurations.

[Grace O'Malley]

You can use also this one: https://yk.github.io/ancestry/

It´s more easy to use and more intuitive than R directly, and you have more options to configurate penalties (or penalization, I don´t know well which word use in English).

When I use it, I use "Adam medium" option, but you can play with different configurations.

Thanks. I am aware of this and should try using it more frequently to understand how it works.

Someone on Anthrogenica posted this ancient model. It was still on there from when I last used it.

Danish_Viking_Average:,0.127131615,0.135299846,0.0 65880231,0.049717154,0.041664462,0.017140923,0.002 259615,0.007082538,0.005003,-0.00014,-0.002598308,0.004046385,-0.008130538,-0.008945385,0.018708538,0.005885,-0.002387077,0.000682308,0.002127385,0.005425692,0. 005701462,0.001112923,0.000151692,0.013551308,-0.000303923
Norwegian_Viking_Average:,0.127671542,0.125037083, 0.069295875,0.059337792,0.04266175,0.02234625,0.00 4416042,0.008663042,0.002053792,-0.004183792,-0.003119208,0.00513925,-0.008938208,-0.010877958,0.020753833,0.011618125,-0.001890667,0.001636375,0.004169,0.007790292,0.005 9895,0.003400417,0.001068125,0.013299917,0.0015767 5
Swedish_Viking_Average:,0.128486176,0.126911382,0. 074547824,0.064714,0.045058088,0.023787676,0.00635 2029,0.010438559,0.004054441,-0.009497706,-0.003863824,0.000740588,-0.002431059,-0.002218176,0.017627676,0.007062324,-0.007228618,0.001184882,0.003852294,0.005693971,0. 008088706,0.0021275,0.000692265,0.011057353,0.0014 79294
Sámi_Average:,0.1044325,-0.036051,0.1057825,0.0792965,-0.01223275,0.008576,0.006169,0.01419175,0.00373275 ,-0.03034225,0.033046,-0.0083175,0.019772,-0.0182695,-0.0065485,-0.0034805,-3.275E-05,-0.00316725,-0.00738475,0.00084425,0.016658,-0.00185475,-0.0016945,0.004097,0.00083825
Gallic_Average:,0.1274818,0.1464394,0.051062,0.038 3078,0.039084,0.0130522,0.001316,0.0038306,0.00535 86,0.0123194,-0.0021108,0.0002398,-0.0112686,-0.0053948,0.0133278,0.005118,-0.009831,0.003826,0.0089998,0.0002002,-0.0037932,0.0062814,0.0029826,-0.0096156,0.0024668
Pictish_Average:,0.1252055,0.135827,0.05845375,0.0 49419,0.03762225,0.018407,-0.000235,0.0049035,0.00598225,0.00469275,-0.0066175,0.008093,-0.01683575,-0.01486325,0.027687,-0.00179,-0.01620025,0.001647,0.00329975,0.0044395,0.00287,0 .004977,-0.00693275,0.016659,0.00266425
Anglo-Saxon_Average:,0.1324615,0.134049875,0.06387475,0. 05692875,0.0411615,0.020568,0.006257,0.006893875,0 .003988125,-0.002414625,-0.004932625,0.004552375,-0.009291375,-0.01245475,0.024243125,0.011237,-0.003080375,0.0015045,0.00520075,0.0066595,0.00533 4375,0.002473125,-0.00164825,0.01664375,-0.000868125

I don't know what they used for the Pictish samples. I would have thought they'd be similar to Gaels but this is what I got with that model.

Distance: 2.414%

Norwegian_Viking_Average: 76.3%
Gallic_Average: 23.7%
Other: 0.0%

Nearest Neighbors
Norwegian_Viking_Average: - 2.693%
Gallic_Average: - 4.569%
Anglo-Saxon_Average: - 5.546%
Pictish_Average: - 5.826%
Danish_Viking_Average: - 6.863%
Swedish_Viking_Average: - 7.221%
Sámi_Average: - 20.326%