Mezcla espanola - Spanish admixture

[E_M81_I3A]

I tested it with a few samples with elevated Amazigh ancestry, it seems not too bad

Target: E_M81_I3a_scaled
Distance: 1.8111% / 0.01811097
52.8 WEST_EUROPEAN
30.8 AMAZIGH
12.6 LEVANTINE
3.8 Sardinian

Some Canarians from the 7 islands :

Target: Spanish_El_Hierro:Can369
Distance: 3.3523% / 0.03352288
56.2 WEST_EUROPEAN
43.8 AMAZIGH

Target: Spanish_La_Gomera:Can332
Distance: 3.7921% / 0.03792130
60.6 WEST_EUROPEAN
38.0 AMAZIGH
1.4 ARABIAN

Target: Spanish_Fuerteventura:Can23
Distance: 3.6407% / 0.03640675
52.0 WEST_EUROPEAN
37.0 AMAZIGH
11.0 Sephardic_Jew

Target: Spanish_Lanzarote:Can48
Distance: 2.6500% / 0.02650046
57.8 WEST_EUROPEAN
36.4 AMAZIGH
3.2 Sardinian
2.6 LEVANTINE

Target: Spanish_Gran_Canaria:Can135
Distance: 4.1680% / 0.04168045
64.8 WEST_EUROPEAN
35.2 AMAZIGH

Target: Spanish_Tenerife:Can290
Distance: 3.3813% / 0.03381257
63.2 WEST_EUROPEAN
32.4 AMAZIGH
4.4 ARABIAN

Target: Spanish_La_Palma:Can26
Distance: 3.7783% / 0.03778301
60.0 WEST_EUROPEAN
29.4 AMAZIGH
10.6 Sardinian


Target: Iberia_Southeast_c.10-16CE:I8146
Distance: 3.3095% / 0.03309503
40.6 WEST_EUROPEAN
36.0 AMAZIGH
14.8 Sardinian
8.6 LEVANTINE

Target: Iberia_Southeast_c.5-8CE:I3575
Distance: 3.3993% / 0.03399274
40.4 WEST_EUROPEAN
32.6 AMAZIGH
18.6 Sephardic_Jew
8.4 Sardinian

Target: ITA_Sardinia_Punic:VIL011
Distance: 2.9432% / 0.02943221
57.4 AMAZIGH
31.8 Sardinian
9.8 ARABIAN
1.0 WEST_EUROPEAN

Target: Levant_LBN_MA_Mixed:SI-41
Distance: 1.7878% / 0.01787781
62.2 WEST_EUROPEAN
20.4 AMAZIGH
9.2 Sardinian
7.6 LEVANTINE
0.6 ARABIAN

Target: ITA_Etruscan_o:RMPR475b
Distance: 2.0408% / 0.02040846
42.4 Sardinian
29.8 AMAZIGH
20.6 WEST_EUROPEAN
7.2 LEVANTINE

Target: ITA_Rome_Imperial:RMPR80
Distance: 1.8869% / 0.01886854
27.6 AMAZIGH
26.0 Sephardic_Jew
23.6 LEVANTINE
16.4 Sardinian
6.4 WEST_EUROPEAN

Target: ITA_Rome_Imperial:RMPR132
Distance: 1.3142% / 0.01314173
49.0 AMAZIGH
24.6 LEVANTINE
18.2 Sardinian
8.2 Sephardic_Jew

[Itrane2000]

Hello
I did the same for Portugal, but I removed Jewish and Corsica as there was too much overlaping.Portugal Admixture.JPG

[Itrane2000]

the main difficulties is to avoid overlaping between the 3 populations: Levantines, Sephardic Jews and Arabians.

[Shadogowah]

Target: Shadogowah_scaled
Distance: 1.4268% / 0.01426793
69.6 WEST_EUROPEAN
16.4 Sardinian
11.2 AMAZIGH
2.8 ARABIAN

[estesiquesabe]

Distance: 2.2266% / 0.02226554
Target: Xabier_scaled
78.0 WEST_EUROPEAN
14.0 Sardinian
8.0 AMAZIGH

[mokordo]

Hola,
testing my model with your coords (yourself, your dad, and mom) is a really good way to find the overlapping.
so my Model is
Amazigh: OK -> as you are the average of your dad and mom
Arabian: NOK -> overlapping with Sephardic Jew and Levantine
Sephardic Jew: NOK -> overlapping with Arabian and Levantine
Levantine : NOK -> overlapping with Arabian and Sephardic
Sardinian: OK
West_European: Globally OK

I have never scored just the average between my mother and my father, neither in simple models I have ocasionally scored less NA than my mother, what´s is quite strange (but not impossible, of course).

My father can score sometimes more in Arab than NA, and being Andalusian quite low.

But you know, all depends on references and how they work together.

Edit:correct text.

[Luso]

Target: Luso_scaled
Distance: 2.1911% / 0.02191073
75.4 WEST_EUROPEAN
13.2 Sephardic_Jew
11.4 AMAZIGH

[Ruderico]

One of the groups includes populations that are very different from one other (West Euro), and since this is the most significant group used in the model it leads to imbalanced and unstable outputs.

If two people from the same region benefit the most from two different sources, for example one from Basques and the other from Provence, the output will forcibly be distinct between them because Basques have virtually no post-roman ancestry which is not the case for Provençals, consequently this will skew the rest of the output, the one who benefits from Basque will probably get a lot more Sephardic, since it's the closest reference to our Mediterranean ancestry. Sardinian also varies a lot but because of different levels of steppe ancestry within the main source.

[mokordo]

One of the groups includes populations that are very different from one other (West Euro), and since this is the most significant group used in the model it leads to imbalanced and unstable outputs.

If two people from the same region benefit the most from two different sources, for example one from Basques and the other from Provence, the output will forcibly be distinct between them because Basques have virtually no post-roman ancestry which is not the case for Provençals, consequently this will skew the rest of the output, the one who benefits from Basque will probably get a lot more Sephardic, since it's the closest reference to our Mediterranean ancestry. Sardinian also varies a lot but because of different levels of steppe ancestry within the main source.

We already know that a model for Iberians based on European references from a non-Iberian country and with Basques as solely Iberian reference, is going to give us a fairly high score in non-European populations, but I don't want to think about any bad intention of the one who posted the model, and I prefer to think that what they want it is not to discriminate well that non-European contribution present in Iberians in its percentage, if not rather in the proportion of different origins of that input.

But I understand perfectly what you mean, and it was my first thought when I see the references used in the model.

But since this is supposed to be a serious forum, away from identity agendas, I prefer to think well.