my final G25 model for a long time

[Kurdquistador]

this time i took 2 Iran N samples that work THE BEST for me and for everybody . these 2 Iran N samples seem to be the closest to CHG and even many Europeans will score a few percent which i will show you below . a Bulgarian individual scores 10% Iran N ! !! . while at the same time i have used Satsurblia CHG instead of Kotias because Satsurblia is older and more "pure" ...Kotias has some EEF and eats a little Iran N . and my fit is also very decent with them


the model : https://justpaste.it/fx6n7

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


my result :

look at these :


Target: Albanian:ALB202
Distance: 4.5794% / 0.04579384
55.8 Balkan_N
20.4 East_European_HG
14.8 Satsurblia_CHG
6.0 Iran_N
3.0 WHG

Target: Albanian:ALB191
Distance: 4.9214% / 0.04921385
57.6 Balkan_N
25.2 East_European_HG
11.0 Satsurblia_CHG
6.0 Iran_N
0.2 WHG




Target: Albanian:AL17
Distance: 4.5458% / 0.04545794
59.4 Balkan_N
17.6 East_European_HG
9.8 Satsurblia_CHG
8.4 Iran_N
2.6 Arabian_HG_Simulated
2.2 WHG



Target: Bulgarian:BulgarianH2
Distance: 4.9191% / 0.04919147
53.6 Balkan_N
21.0 East_European_HG
10.6 Iran_N
7.8 Satsurblia_CHG
4.6 WHG
2.4 Arabian_HG_Simulated


Target: Bulgarian:BulgarianE2
Distance: 4.5647% / 0.04564662
51.2 Balkan_N
29.0 East_European_HG
10.0 Satsurblia_CHG
6.0 Iran_N
3.8 WHG

 

[Kurdquistador]

Target: Romanian:G421
Distance: 5.1372% / 0.05137160
56.8 Balkan_N
26.2 East_European_HG
7.4 Iran_N
6.0 Satsurblia_CHG
3.6 WHG



Target: Romanian:A374
Distance: 4.6532% / 0.04653192
50.2 Balkan_N
30.2 East_European_HG
9.0 Satsurblia_CHG
6.4 Iran_N
4.2 WHG

 

[Kurdquistador]

when you use the south AASI (Hakkipikki) and take out the northwest AASI then this iranian shirazi individual scores like this . (with nw-aasi he/she gets 4.4)

Target: Iranian_Persian_Shiraz:SHII94
Distance: 2.3698% / 0.02369815
35.0 Iran_N
26.8 Balkan_N
13.2 Satsurblia_CHG
8.8 East_European_HG
7.6 Arabian_HG_Simulated
5.4 AASI
2.8 Russia_DevilsCave_N.SG
0.4 Taforalt


Baloch average sample with hakkipikki aasi

Target: Balochi
Distance: 2.0046% / 0.02004596
62.2 Iran_N
14.6 Balkan_N
13.8 East_European_HG
8.4 AASI
0.8 Taforalt
0.2 WHG

 

[Kurdquistador]

look at these 2 Baloch individuals . shout out to the fuckheads who are darkwashing and exotifying Iran Neolithic by saying "look at balochs how they look" . well you fucking pieces of shits Balochs have significant AASI ....PLUS they inhabit a very hot climate with lots of sun . these 2 things will turn many balochs into pajeetoids . AASI is VERY dark and also non caucasian racially . imagine you had 8%+ even 10%+ AASI and are cooking around under the sun . of course you will look like that


Target: Balochi:HGDP00098
Distance: 3.0856% / 0.03085632
53.8 Iran_N
13.8 Balkan_N
12.2 East_European_HG
12.0 AASI
6.2 Satsurblia_CHG
1.2 Taforalt
0.8 SSA



Target: Balochi:HGDP00090
Distance: 2.2635% / 0.02263454
60.6 Iran_N
14.2 AASI
14.0 East_European_HG
11.0 Balkan_N
0.2 SSA

 

[Ben Dover]

what happened to your claims about Iran_N being exaggerated and CHG being underestimated in most models. Now that you use satsurblia it exaggerates Iran_N and reduces CHG even more than it usually happens with different sampling. And satsurblia is not that relevant for us anyway. Kotia is more recent, higher coverage and it is still very similar to satsurblia at the end of the day

 

[Ben Dover]

and why would you use Iran_N samples that are most similar to CHG if your aim is to prevent overfit and unmasking true CHG

 

[Kurdquistador]

what happened to your claims about Iran_N being exaggerated and CHG being underestimated in most models. Now that you use satsurblia it exaggerates Iran_N and reduces CHG even more than it usually happens with different sampling. And satsurblia is not that relevant for us anyway. Kotia is more recent, higher coverage and it is still very similar to satsurblia at the end of the day

this time i did the reverse . i have used 2 Iran N that are more like CHG

you can use Kotias but i prefer Satsu because Kotias is younger than Satsu and has some small drifts compared to Satsu

and just like the previous models / thread i made was supposed to show hidden CHG .....this model is supposed to show hidden Iran N

 

[Kurdquistador]

and why would you use Iran_N samples that are most similar to CHG if your aim is to prevent overfit and unmasking true CHG

my aim this time was to minimize the difference of Iran N and CHG

 

[Ben Dover]

have you tried modeling satsurblia with kotia as a base to see how satsurblia shifts compared to kotia. whatever is younger is more relevant in a general sense. but it makes no sense to first try to uncover hodden CHG and then to uncover hidden Iran_N. it does not work that way. it is either one or the other. Similarly it would not make sense to uncover hidden natufian from ANF and vice versa

 

[idklolimo]

 

[Kurdquistador]

have you tried modeling satsurblia with kotia as a base to see how satsurblia shifts compared to kotia. whatever is younger is more relevant in a general sense. but it makes no sense to first try to uncover hodden CHG and then to uncover hidden Iran_N. it does not work that way. it is either one or the other. Similarly it would not make sense to uncover hidden natufian from ANF and vice versa

satsurblia doesnt get modelled because it is older i guess

Target: Satsurblia_CHG
Distance: 3.2997% / 0.03299736
100.0 Georgia_Kotias.SG


while Kotias :


Target: Georgia_Kotias.SG
Distance: 2.4692% / 0.02469195
90.4 Satsurblia_CHG
4.0 Balkan_N
4.0 Iran_N
0.8 East_European_HG
0.4 Taforalt
0.2 AASI
0.2 SSA



no it is not one or the other they overlap thats the thing . you can either choose some Iran N that is less like CHG to show more CHG or you can use Iran N that is more like CHG which will have the reverse effect

 

[Ben Dover]

no it is not one or the other they overlap thats the thing . you can either choose some Iran N that is less like CHG to show more CHG or you can use Iran N that is more like CHG which will have the reverse effect

it does not make sense because realistically you cant go from being lets say 33% CHG to just 24% or from 26% Iran_N to just 29% Iran_N and anything in between. You cant also change those scores like a sock everytime you make a new model

 

[Kurdquistador]

it does not make sense because realistically you cant go from being lets say 33% CHG to just 24% or from 26% Iran_N to just 29% Iran_N and anything in between. You cant also change those scores like a sock everytime you make a new model

it always depends which pops you use . if they eat up each other or not

i like this model . you dont seem to like it unfortunately .

 

[Ben Dover]

are you 10% CHG or 18% CHG per your two latest models which uncover masked CHG and Iran_N respectively

 

[Kurdquistador]

are you 10% CHG or 18% CHG per your two latest models which uncover masked CHG and Iran_N respectively

both

i know it sounds odd and i am not sure how to explain it lol

 

[Kurdquistador]

funnily enough Serbia N gives a slightly better fit than Barcin LN

Target: fuckeeeeer_scaled
Distance: 3.4311% / 0.03431066

37.8 Turkey_Barcin_LN.SG
37.2 Iran_N
9.2 Satsurblia_CHG
7.8 Arabian_HG_Simulated
6.2 East_European_HG
1.8 Russia_DevilsCave_N.SG

 

[KARABOĞA]

what does satsurblia refer to?

 

[KARABOĞA]

what does satsurblia refer to?

View attachment 15866

he chg yazıyomuş şimdi gördüm

 

[Ben Dover]

what does satsurblia refer to?

View attachment 15866

can you use kotias.SG instead of Satsurblia for CHG and Natufian instead of arabian_HG_sim