Iberian+British Islands+French+Italian G25 customized modern references scaled model

[gixajo]

The idea is to test averages of these four references until each person from these places scores as high as possible in their origin. I will try to improve the models, when I can.

With many of us it will be practically impossible to avoid overlapping, but at least try to discriminate each origin as best as possible.

SCALED

https://vahaduo.github.io/vahaduo/

ADD DIST COL : NO or 0.25X (at your choice)


British_Islands,0.13165413,0.13707704,0.06136654,0 .04485214,0.03995564,0.01709027,0.00374873,0.00575 732,0.00510017,0.00480905,-0.00518153,0.00578464,-0.01307829,-0.01129563,0.02183706,0.00395673,-0.01089574,0.00277469,0.00292593,0.00235050,0.0050 8376,0.00321609,-0.00117769,0.01363141,0.00036479
Iberian,0.1125447,0.1463265,0.0405907,-0.0012957,0.0446156,-0.0023888,-0.0026005,0.0017872,0.0249589,0.0323751,-0.0025929,0.0076949,-0.0141516,-0.0123163,0.0108789,-0.0007383,-0.0044854,-0.0007214,-0.0029596,-0.0014927,0.0035598,-0.0015829,-0.0018367,-0.0031920,0.0008707
Italian,0.1145311,0.147316,0.0116278,-0.0249723,0.0244378,-0.0084909,0.0006091,-0.0000722,0.0079627,0.0228155,-0.0000746,0.0044706,-0.008622,-0.0017359,-0.0041116,-0.0028368,0.0007306,0.0004706,0.0029821,-0.0038168,-0.0016601,0.0015952,-0.0006129,0.0018923,-0.0004877
French,0.1274933,0.1429826,0.0495027,0.0193734,0.0 428681,0.0072217,0.0015815,0.0033606,0.0139183,0.0 205586,-0.0051245,0.0064464,-0.0138497,-0.0098263,0.0112950,0.0030423,-0.0042234,0.0021441,0.0013707,-0.0004168,0.0041569,0.0022650,-0.0024467,0.0031867,-0.0006697


British_Islands,=(English_Cornwall+English)+(Irish +Scottish+Welsh)
Iberian,=Iberian:Center,+Iberian:NorthWest_Iberia, +Iberian:North_Iberia,+Iberian:NorthEast_Iberia,+I berian:SouthWest_Iberia,+Iberian:SouthEast_Iberia,
Italian=Italian:Center,+Italian:South_Italian:Main land,+Italian:South_Italian:Sicilian,+NorthWest_It alian_B,+Italian:NorthEast_Italian,
French=French:Center,+French:France_Nord-Ouest,+French:France:France_Nord-Est,+French:France_Sud-Est,+French:France_Sud-Ouest,



Iberian:Center=CastillaYLeón+CastillaLaMancha
Iberian:NorthWest_Iberia=Portuguese,+Spanish_Galic ia,
Iberian:North_Iberia=Spanish_Navarra,+Spanish_Pais _Vasco,+Spanish_Cantabria,
Iberian:NorthEast_Iberia=Spanish_Valencia,+Spanish _Cataluna,
Iberian:SouthWest_Iberia,=Spanish_Andalucia,+Spani sh_Extremadura,+Portuguese,
Iberian:SouthEast_Iberia=Spanish_Murcia,+Spanish_A ndalucia,+Spanish_Castilla_La_Mancha


Italian:Center= Italian_Marche,+Italian_Lazio,+Italian_Tuscany,+It alian_Umbria
Italian:Southern_Mainland_Italian,=Italian_Basilic ata,+Italian_Campania,+Italian_Apulia,+Italian_Cal abria,
Italian:Sicilia=Sicilia_East+Sicilia_West
Italian:NorthWest_Italian_B,=Italian_Bergamo,+Ital ian_Lombardy,+Italian_Liguria,+Italian_Piedmont,
Italian:Northeast_Italian,=Italian_Veneto,+Italian _Trentino-Alto-Adige,+Italian_Northeast,


French:Center=Paris+Auvergne
French:France_Nord-Ouest=French_Brittany,+French_Pas-de-Calais,+French_Seine-Maritime,
French:France_Nord-Est=French_Nord+French_Alsace
French:France_Sud-Est=French_Provence,+French_Occitanie,
French:France_Sud-Ouest=French_Basque+French_South

[mitalit]

Target: Xabier_scaled
Distance: 2.5662% / 0.02566216
65.2 Iberian
20.6 British_Islands
14.2 Italian

Target: Xabier_scaled
Distance: 2.5856% / 0.02585603 | ADC: 0.25x
66.4 Iberian
13.2 British_Islands
10.4 French
10.0 Italian

[gixajo]

Target: gixajo_scaled
Distance: 1.7604% / 0.01760397
79.2 Iberian
20.8 French

Target: gixajo_dad_scaled
Distance: 2.0863% / 0.02086306
88.8 Iberian
11.2 Italian

Target: gixajo_mom_scaled
Distance: 2.2316% / 0.02231621
80.4 Iberian
19.6 British_Islands
*****************************************



Target: gixajo_scaled
Distance: 1.7613% / 0.01761278 | ADC: 0.25x
80.8 Iberian
19.2 French

Target: gixajo_dad_scaled
Distance: 2.0880% / 0.02088028 | ADC: 0.25x
90.4 Iberian
9.6 Italian

Target: gixajo_mom_scaled
Distance: 2.2365% / 0.02236544 | ADC: 0.25x
80.4 Iberian
17.0 British_Islands
2.6 French

[oszkar07]

Target: Oszkar
Distance: 3.6323% / 0.03632276 | ADC: 0.25x
88.4 British_Islands
11.6 Italian

Target: Oszkar
Distance: 3.6292% / 0.03629231
87.0 British_Islands
13.0 Italian

[gixajo]

[B][B]T,,,,,n

Nada raro, no?

I am interested in people who has mixed origins from these areas, to see if the model discriminates well their origins.

And natives from any of those zones, of course.

[gixajo]

Target: Oszkar
Distance: 3.6323% / 0.03632276 | ADC: 0.25x
88.4 British_Islands
11.6 Italian

Target: Oszkar
Distance: 3.6292% / 0.03629231
87.0 British_Islands
13.0 Italian

And adding this customized one I´ve made time ago?


Hungarian:B,0.126710,0.125490,0.044271,0.034419,0. 034321,0.011358,0.006558,0.006420,0.004916,-0.005837,-0.002537,-0.001960,0.003476,0.008790,-0.004186,0.001321,0.004641,-0.000695,0.002875,-0.000161,-0.002889,0.000394,0.004839,0.002371,0.000190

[oszkar07]

And adding this customized one I´ve made time ago?


Hungarian:B,0.126710,0.125490,0.044271,0.034419,0. 034321,0.011358,0.006558,0.006420,0.004916,-0.005837,-0.002537,-0.001960,0.003476,0.008790,-0.004186,0.001321,0.004641,-0.000695,0.002875,-0.000161,-0.002889,0.000394,0.004839,0.002371,0.000190

Works reasonably good , captures my recent known ancestry ...but the samples you can see in my signature below give me better distance but your samples are closer to 50/50 split.

Target: Oszkar
Distance: 2.8330% / 0.02833037 | ADC: 0.25x
50.8 Hungarian
49.2 British_Islands

Target: Oszkar
Distance: 2.8330% / 0.02833037
50.8 Hungarian
49.2 British_Islands

[Ibericus]

Distance: 2.3256% / 0.02325567
66.0 Iberian
20.8 Italian
10.2 French
3.0 British_Islands

[Grace O'Malley]

The only pull I get outside of the Isles is Scandinavia so unless they are added I get this.

Target: Grace_scaled
Distance: 2.1056% / 0.02105640 | ADC: 0.25x
100.0 British_Islands
0.0 French
0.0 Iberian
0.0 Italian

Target: Grace_scaled
Distance: 2.1056% / 0.02105640
100.0 British_Islands
0.0 French
0.0 Iberian
0.0 Italian

[gixajo]

Works reasonably good , captures my recent known ancestry ...but the samples you can see in my signature below give me better distance but your samples are closer to 50/50 split.

s

Of course, your signature samples are individual samples, and mines are averages, we will more often obtain better distances with some individual samples than with averages.

Anyway your distance is under 3 , so quite well if your main recent ancestries are well discriminated.:thumb001:

Edit: Take into account the stability in distances, you have the same distance without penalization and 0.25x.

[gixajo]

The only pull I get outside of the Isles is Scandinavia so unless they are added I get this.

n

British Islands reference is not new, only Italian and French ones are very recent and never used, and Iberian one was used once before making this model.

The creator of the G25 seems to take special care of references from northwestern Europe, and especially those of the British Isles.

[Pedro Ruben]

Target: Pedro_scaled
Distance: 3.5301% / 0.03530101
87.2 Iberian
12.2 Italian
0.6 British_Islands

[Defcon2]

Target: Juan_scaled
Distance: 3.2509% / 0.03250908
93.0 Iberian
7.0 Italian

[Celestia]

Target: Celestia_scaled
Distance: 4.8355% / 0.04835479 | ADC: 0.25x
68.2 British_Islands
31.8 Iberian

[JamesBond007]

British Islands reference is not new, only Italian and French ones are very recent and never used, and Iberian one was used once before making this model.

The creator of the G25 seems to take special care of references from northwestern Europe, and especially those of the British Isles.

The only pull get outside of Britain are Dutch and Norwegian, with G25, so unless they are added my distance is this below. Special care ? Unless they are academic samples like the ancient England Saxon Hinxton etc.. then they are trash in general because you can't trust self-reported ancestry but an even bigger problem is that the ethnic classification of the Global25 samples is too fine-grained, especially for modern European samples.

Target: Kevin_scaled
Distance: 3.1620% / 0.03162041
100.0 British_Islands

[alnortedelsur]

Target: Alnortedelsur_scaled
Distance: 6.9692% / 0.06969197
69.0 Iberian
31.0 Italian

Target: Alnortedelsur_scaled
Distance: 6.9694% / 0.06969426 | ADC: 0.25x
70.0 Iberian
30.0 Italian

Distance to: Alnortedelsur_scaled
0.07158335 Iberian
0.07855861 Italian
0.08673351 French
0.10686584 British_Islands

[RyoHazuki]

Target: Ryo_scaled
Distance: 3.0259% / 0.03025906
91.0 British_Islands
6.2 Iberian
2.8 Italian

[calxpal]

Target: cpalscaled
Distance: 3.7494% / 0.03749434
81.2 British_Islands
18.8 Italian

[grabielx]

Distance to:
0.03153696 Italian
0.04809817 Iberian
0.05692837 French
0.08819228 British_Islands

Target:
Distance: 2.5958% / 0.02595785
73.8 Italian
26.2 French

[gixajo]

The only pull get outside of Britain are Dutch and Norwegian, with G25, so unless they are added my distance is this below. Special care ? Unless they are academic samples like the ancient England Saxon Hinxton etc.. then they are trash in general because you can't trust self-reported ancestry but an even bigger problem is that the ethnic classification of the Global25 samples is too fine-grained, especially for modern European samples.

Target: Kevin_scaled
Distance: 3.1620% / 0.03162041
100.0 British_Islands

All that is very good, but apart from that, there are more things, and if you have had your eyes open you should know what I mean.

There are three recidivist(?) "effects" that I find quite strange, and that I have seemed to observe in the G25 and in the models that the members usually make. And at least one of these effects is present clearly in the results of this model, and this effect will explain why I say some references seem to have been "specially treated".

I am not speaking about genetics themselves, but about the internal functioning of this tool and how references are made to work and interactuate when they are run with other references.

[gixajo]

Distance to:
0.03153696 Italian
0.04809817 Iberian
0.05692837 French
0.08819228 British_Islands

Target:
Distance: 2.5958% / 0.02595785
73.8 Italian
26.2 French

Can I know where aproximately is your Italian ancestry from?

[RyoHazuki]

All that is very good, but apart from that, there are more things, and if you have had your eyes open you should know what I mean.

There are three recidivist(?) "effects" that I find quite strange, and that I have seemed to observe in the G25 and in the models that the members usually make. And at least one of these effects is present clearly in the results of this model, and this effect will explain why I say some references seem to have been "specially treated".

I am not speaking about genetics themselves, but about the internal functioning of this tool and how references are made to work and interactuate when they are run with other references.

Could you please go into more specifics on the effects of this tool? I'm curious now.

[celticdragongod]

Distance to: CDG_scaled
0.01748371 British_Islands
0.04921616 French
0.08031275 Iberian
0.10743691 Italian

Target: CDG_scaled
Distance: 1.7484% / 0.01748371
100.0 British_Islands

[gixajo]

Could you please go into more specifics on the effects of this tool? I'm curious now.

No, I could be more specific but I don´t want to be more specific.;)

But I gave already a clue, in this model we can see clearly one of those curious "effects".

[RyoHazuki]

No, I could be more specific but I don´t want to be more specific.;)

But I gave already a clue, in this model we can see clearly one of those curious "effects".

https://i.redd.it/d51a38p36b221.jpg

[Rocinante]

0.03975161 Iberian
0.04485290 Italian
0.05705629 French

Distance: 3.2976% / 0.03297584
57.8 Iberian
42.2 Italian

[gixajo]

[IM

I'm not David Lynch talking about a movie of his.

I'm someone who has nothing to do with those who have created the references and the tools we use to run, so if I see something strange it may simply be that I do not understand well how it works all this, and not a bug of the process itself, or something intentional.

And since I do not know why what I have observed happens, although I have some idea of ​​what it could be due to, since I have no certainty, I better shut up.

[Damião de Góis]

Target: Gil_Vicente
Distance: 2.5130% / 0.02513038
77.4 Iberian
14.8 Italian
7.8 British_Islands

Target: Gil_Vicente
Distance: 2.5207% / 0.02520749 | ADC: 0.25x
82.8 Iberian
11.8 Italian
5.4 British_Islands

[Coastal Elite]

Target: DacoCeltic_scaled
Distance: 2.2290% / 0.02228980 | ADC: 0.25x
65.8 British_Islands
34.2 Italian

Target: DacoCeltic_scaled
Distance: 2.2273% / 0.02227283
65.0 British_Islands
35.0 Italian

[gixajo]

Target: DacoCeltic_scaled
n

An adding this one?

Vlach=Romanian+Moldavian


Vlach,0.1242270,0.130858,0.037156,0.0144255,0.0282 41,0.0057105,0.005408,0.006261,0.000137,0.0020645,-0.0025935,-0.00109,0.002249,0.0110690,-0.011775,-0.0021085,0.005019,-0.000919,0.003392,-0.0021025,-0.0053775,-0.001126,0.003622,0.000366,-0.000794

[Beowulf]

Target: Ivorix
Distance: 1.3097% / 0.01309726
48.8 Italian
36.0 Iberian
15.2 British_Islands






Target: mama
Distance: 1.8382% / 0.01838174
34.6 French
34.2 Italian
31.2 Iberian

Target: papa
Distance: 0.5520% / 0.00551975
51.0 Iberian
29.0 Italian
20.0 British_Islands



Target: Abuela
Distance: 2.3088% / 0.02308828
56.6 Iberian
24.4 British_Islands
19.0 Italian

Target: Abuelo
Distance: 0.7245% / 0.00724499
56.2 French
25.4 Italian
18.4 Iberian

[vader]

Target: Luso_scaled
Distance: 3.0317% / 0.03031686
92.2 Iberian
7.8 Italian

Target: Luso_dad_scaled
Distance: 2.0386% / 0.02038635
100.0 Iberian

Target: Luso_mom_scaled
Distance: 3.1329% / 0.03132873
86.4 Iberian
13.6 Italian

[Beowulf]

Paternal Grandpa:

Target: Abuelo_Bartolome_Scaled
Distance: 1.4484% / 0.01448369
64.8 Iberian
18.4 Italian
9.4 British_Islands
7.4 French

Paternal Grandma:



Target: Abuela_Josefina_Scaled
Distance: 0.6694% / 0.00669430
68.2 Iberian
31.8 Italian

Great Grandma:

Target: Bisabuela_Cristobalina_Scaled
Distance: 0.5512% / 0.00551156
71.6 Iberian
28.4 Italian

[JerryS.]

Target: JerryS_scaled
Distance: 1.9697% / 0.01969724
89.4 British_Islands
10.6 Italian