Dodecad k12b south, west and central asian results

[Demhat]

Totally agree. I wouldn't call him an idiot. Clearly he's not a scientist if he's relying exclusively on G25. I would just call him a misinformed amateur who doesn't understand the issues with Davidski's G25.

The Hurrian he's referring to I think is Haji-Firuz-IA ! and the the Archemenid I believe is Hasanlu-IA. Clearly the G25 is giving totally off results. Davidski's G25 is a mish mosh of modern and ancient samples. Obviously the moderns he has in the run to generate coordinates are having all sorts of effects on the test samples.

Anyways if he was a scientist he would use qpAdm and if he had he would have seen the following for Haji-Firuz-IA (courtesy EurasianDNA)





So what the G25 shows is irrelevant. Haji-Firuz-IA is actually 67.5% Iran-Chl + 32.5% Central Asian Sarmatian related with a p-value of 0.78 (excellent p-value, model passing p-values used in scientific papers is 0.05)

Wait a minute now is it a real Achaemenid sample or not? From where and which timeframe is this Hasanlu Iron Age sample exactly. Haji Firuz might be genetically very close to Medes.

I don't call him an idiot he is just not well informed and doesn't know the flaws and how these calculators work. If I had no other proxy samples than a Chimp and East Asian. The calculator would obviously pick up East Asian for me, because it's the next closest thing that is there.

[Eline]

Wait a minute now is it a real Achaemenid sample or not?

I think people here are confused and speak about the different samples and topics.

There is no Achaemenid sample at the PCA you have seen. That PCA is from a recent study about the Iranian people: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6759149/

[Zoro]

Wait a minute now is it a real Achaemenid sample or not? From where and which timeframe is this Hasanlu Iron Age sample exactly. Haji Firuz might be genetically very close to Medes.

I don't call him an idiot he is just not well informed and doesn't know the flaws and how these calculators work. If I had no other proxy samples than a Chimp and East Asian. The calculator would obviously pick up East Asian for me, because it's the next closest thing that is there.

Hasanlu-IA is a 2750 year old sample from Hasanlu Iran in Zagros. So about 250 years before Archemenids It's R1b1a1a2a2

[Demhat]

So Iranians are more steppe than anticipated.

Seems so but the thing people need to understand and always remember. Inside this ~30% Steppe hides around 15-20% Iranian_Plateau Neolithic, Neolithic Caucasus related DNA. It was a bi directional geneflow. Those dudes who came from the Steppe (especially the later Iranic branches) were at least 50% of Iranian Plateau-Caucasus ancestry if not even more. That is why we barely show outside admixture, because allot of it is literally back migration.

[Zoro]

Haji-Firuz-IA is from Haji Firuz Iran and is 2900 years old. Y:R1b1a1a2a2 Mt: K1a17a

[Demhat]

Hasanlu-IA is a 2750 year old sample from Hasanlu Iran in Zagros. So about 250 years before Archemenids It's R1b1a1a2a2

So it might actually be a Mede, timeframe fits. But it's slightly before the foundation of the Median Empire

[Demhat]

Haji-Firuz-IA is from Haji Firuz Iran and is 2900 years old. Y:R1b1a1a2a2 Mt: K1a17a

These guys were either already Medes or Hurrians-Manneaens/Gutians/Kassites who were about to become the Medes. Either way they possible represent well what the early Median speakers genetically looked like.

Still would be good to have samples from the actual period of the Median Empire and the Achaemenid Dynasty.

[Eline]

So it might actually be a Mede, timeframe fits. But it's slightly before the foundation of the Median Empire

Yezidi Kurds from Iraq have much more R1b than R1a. Non-Yezidi Kurds have more R1a

Yezidi Kurds have also a lot Maykop/Kura-Araxes Y-DNA hg. 'L'.

I think that R1b is connected to the ancient proto-West Iranian Gutians who were ancestral to the Mitanni, Medes.

[Demhat]

Yezidi Kurds from Iraq have much more R1b than R1a. Non-Yezidi Kurds have more R1a

Yezidi Kurds have also a lot Maykop/Kura-Araxes Y-DNA hg. 'L'.

I think that R1b is connected to the ancient Gutians who were ancestral to the Medes.

The Medes were literally a confederation of Iranic and non Iranic tribes who spoke the Median language. Cyrus literally called all of West Media Gutium (land of the Gutis).

The Persian were literally a synthesis of Iranic and non Iranic tribes such as the Elamites and Kassites which formed the Persians.

The very Medes and Persian existence is down to these synthesis of Iranic and non Iranic speakers. We should never forget that.

Before these Iranic tribes mixed with the other locals they literally called themselves just "Aryans".

[Zoro]

Seems so but the thing people need to understand and always remember. Inside this ~30% Steppe hides around 15-20% Iranian_Plateau Neolithic, Neolithic Caucasus related DNA. It was a bi directional geneflow. Those dudes who came from the Steppe (especially the later Iranic branches) were at least 50% of Iranian Plateau-Caucasus ancestry if not even more. That is why we barely show outside admixture, because allot of it is literally back migration.

Very good point and that's one thing that confounds ADMIXTURE or G25 results because they don't use outgroups to zero out common shared ancestry. A big example is Europeans are shown closer to Sarmatians or Scythians in ADMIXTURE or G25 not because they had alot of admixture from those groups but because steppe groups and Europeans share alot of ancient EHG admixture whereas the basal Eurasian in W. Asians puts them further away from Steppe.

qpAdm on the other hand DOES use OUTGROUPS and much less vulnerable to this issue. Here's Harji-Firuz-IA using the following ancient OUTGROUPs to zero out ancient shared alleles between Haji-Firuz-IA and C. Asian Sarmatian

Mbuti.DG
Russia_Ust_Ishim.DG
Onge_1000G
Russia_MA1_HG.SG
Chinese_Han_1000G
Karitiana.DG
Anatolia_N
Russia_Shamanka_Eneolithic
Georgia_Kotias.SG

numthreads: 8
## qpAdm version: 634
seed: 1215982657

left pops:
Iran_IA_HajjiFiruz
Iran_C_HajjiFiruz
Kazakhstan_Sarmatian.SG

right pops:
Mbuti.DG
Russia_Ust_Ishim.DG
Onge_1000G
Russia_MA1_HG.SG
Chinese_Han_1000G
Karitiana.DG
Anatolia_N
Russia_Shamanka_Eneolithic
Georgia_Kotias.SG

0 Iran_IA_HajjiFiruz 1
1 Iran_C_HajjiFiruz 4
2 Kazakhstan_Sarmatian.SG 2
3 Mbuti.DG 4
4 Russia_Ust_Ishim.DG 1
5 Onge_1000G 6
6 Russia_MA1_HG.SG 1
7 Chinese_Han_1000G 211
8 Karitiana.DG 3
9 Anatolia_N 29
10 Russia_Shamanka_Eneolithic 10
11 Georgia_Kotias.SG 1
jackknife block size: 0.050
snps: 709309 indivs: 273
number of blocks for block jackknife: 711
## ncols: 709309
coverage: Iran_IA_HajjiFiruz 292807
coverage: Iran_C_HajjiFiruz 624648
coverage: Kazakhstan_Sarmatian.SG 474638
coverage: Mbuti.DG 690580
coverage: Russia_Ust_Ishim.DG 654138
coverage: Onge_1000G 704593
coverage: Russia_MA1_HG.SG 493659
coverage: Chinese_Han_1000G 709306
coverage: Karitiana.DG 690550
coverage: Anatolia_N 709269
coverage: Russia_Shamanka_Eneolithic 709182
coverage: Georgia_Kotias.SG 708661
dof (jackknife): 639.275
numsnps used: 154924
codimension 1
f4info:
f4rank: 1 dof: 7 chisq: 7.220 tail: 0.406371316 dofdiff: 9 chisqdiff: -7.220 taildiff: 1
B:
scale 1.000
Russia_Ust_Ishim.DG 0.182
Onge_1000G 0.323
Russia_MA1_HG.SG 1.249
Chinese_Han_1000G 0.915
Karitiana.DG 1.548
Anatolia_N -1.191
Russia_Shamanka_Eneolithic 1.269
Georgia_Kotias.SG -0.207
A:
scale 734.456
Iran_C_HajjiFiruz -0.547
Kazakhstan_Sarmatian.SG 1.304


full rank
f4info:
f4rank: 2 dof: 0 chisq: 0.000 tail: 1 dofdiff: 7 chisqdiff: 7.220 taildiff: 0.406371316
B:
scale 1.000 1.000
Russia_Ust_Ishim.DG 0.199 0.121
Onge_1000G 0.324 0.590
Russia_MA1_HG.SG 1.234 -0.251
Chinese_Han_1000G 0.892 1.134
Karitiana.DG 1.545 0.215
Anatolia_N -1.181 0.885
Russia_Shamanka_Eneolithic 1.298 -0.464
Georgia_Kotias.SG -0.262 2.290
A:
scale 748.617 2653.477
Iran_C_HajjiFiruz -0.655 -1.132
Kazakhstan_Sarmatian.SG 1.254 -0.848


best coefficients: 0.704 0.296
Jackknife mean: 0.704624134 0.295375866
std. errors: 0.068 0.068

error covariance (* 1000000)
4689 -4689
-4689 4689


summ: Iran_IA_HajjiFiruz 2 0.406371 0.705 0.295 4689 -4689 4689

fixed pat wt dof chisq tail prob
00 0 7 7.220 0.406371 0.704 0.296
01 1 8 24.876 0.00163144 1.000 0.000
10 1 8 83.531 9.49035e-15 0.000 1.000
best pat: 00 0.406371 - -
best pat: 01 0.00163144 chi(nested): 17.657 p-value for nested model: 2.64607e-05

coeffs: 0.704 0.296

## dscore:: f_4(Base, Fit, Rbase, right2)
## genstat:: f_4(Base, Fit, right1, right2)

details: Iran_C_HajjiFiruz Russia_Ust_Ishim.DG -0.000301 -0.479523
details: Kazakhstan_Sarmatian.SG Russia_Ust_Ishim.DG 0.000231 0.321952
dscore: Russia_Ust_Ishim.DG f4: -0.000144 Z: -0.240060

details: Iran_C_HajjiFiruz Onge_1000G -0.000612 -1.057869
details: Kazakhstan_Sarmatian.SG Onge_1000G 0.000288 0.443516
dscore: Onge_1000G f4: -0.000346 Z: -0.630565

details: Iran_C_HajjiFiruz Russia_MA1_HG.SG -0.001029 -1.354396
details: Kazakhstan_Sarmatian.SG Russia_MA1_HG.SG 0.002093 2.330573
dscore: Russia_MA1_HG.SG f4: -0.000106 Z: -0.145244

details: Iran_C_HajjiFiruz Chinese_Han_1000G -0.001320 -2.749701
details: Kazakhstan_Sarmatian.SG Chinese_Han_1000G 0.001089 1.979980
dscore: Chinese_Han_1000G f4: -0.000608 Z: -1.332328

details: Iran_C_HajjiFiruz Karitiana.DG -0.001525 -2.377830
details: Kazakhstan_Sarmatian.SG Karitiana.DG 0.002446 3.271362
dscore: Karitiana.DG f4: -0.000352 Z: -0.566976

details: Iran_C_HajjiFiruz Anatolia_N 0.000583 1.180980
details: Kazakhstan_Sarmatian.SG Anatolia_N -0.002324 -4.105589
dscore: Anatolia_N f4: -0.000276 Z: -0.583845

details: Iran_C_HajjiFiruz Russia_Shamanka_Eneolithic -0.001013 -1.939311
details: Kazakhstan_Sarmatian.SG Russia_Shamanka_Eneolithic 0.002256 3.544665
dscore: Russia_Shamanka_Eneolithic f4: -0.000047 Z: -0.092013

details: Iran_C_HajjiFiruz Georgia_Kotias.SG -0.000884 -1.152117
details: Kazakhstan_Sarmatian.SG Georgia_Kotias.SG -0.001299 -1.522396
dscore: Georgia_Kotias.SG f4: -0.001006 Z: -1.382609

gendstat: Mbuti.DG Russia_Ust_Ishim.DG -0.240
gendstat: Mbuti.DG Onge_1000G -0.631
gendstat: Mbuti.DG Russia_MA1_HG.SG -0.145
gendstat: Mbuti.DG Chinese_Han_1000G -1.332
gendstat: Mbuti.DG Karitiana.DG -0.567
gendstat: Mbuti.DG Anatolia_N -0.584
gendstat: Mbuti.DG Russia_Shamanka_Eneolithic -0.092
gendstat: Mbuti.DG Georgia_Kotias.SG -1.383
gendstat: Russia_Ust_Ishim.DG Onge_1000G -0.341
gendstat: Russia_Ust_Ishim.DG Russia_MA1_HG.SG 0.049
gendstat: Russia_Ust_Ishim.DG Chinese_Han_1000G -0.902
gendstat: Russia_Ust_Ishim.DG Karitiana.DG -0.327
gendstat: Russia_Ust_Ishim.DG Anatolia_N -0.240
gendstat: Russia_Ust_Ishim.DG Russia_Shamanka_Eneolithic 0.172
gendstat: Russia_Ust_Ishim.DG Georgia_Kotias.SG -1.110
gendstat: Onge_1000G Russia_MA1_HG.SG 0.315
gendstat: Onge_1000G Chinese_Han_1000G -0.630
gendstat: Onge_1000G Karitiana.DG -0.010
gendstat: Onge_1000G Anatolia_N 0.139
gendstat: Onge_1000G Russia_Shamanka_Eneolithic 0.629
gendstat: Onge_1000G Georgia_Kotias.SG -0.898
gendstat: Russia_MA1_HG.SG Chinese_Han_1000G -0.751
gendstat: Russia_MA1_HG.SG Karitiana.DG -0.342
gendstat: Russia_MA1_HG.SG Anatolia_N -0.243
gendstat: Russia_MA1_HG.SG Russia_Shamanka_Eneolithic 0.085
gendstat: Russia_MA1_HG.SG Georgia_Kotias.SG -1.002
gendstat: Chinese_Han_1000G Karitiana.DG 0.614
gendstat: Chinese_Han_1000G Anatolia_N 0.822
gendstat: Chinese_Han_1000G Russia_Shamanka_Eneolithic 2.061
gendstat: Chinese_Han_1000G Georgia_Kotias.SG -0.579
gendstat: Karitiana.DG Anatolia_N 0.140
gendstat: Karitiana.DG Russia_Shamanka_Eneolithic 0.651
gendstat: Karitiana.DG Georgia_Kotias.SG -0.844
gendstat: Anatolia_N Russia_Shamanka_Eneolithic 0.488
gendstat: Anatolia_N Georgia_Kotias.SG -1.092
gendstat: Russia_Shamanka_Eneolithic Georgia_Kotias.SG -1.309

## end of run

So even using all those OUTGROUPS to zero out ancient shared alleles between Haji-Firuz and C. Asian Sarmatian we still get


70.4% +/-7% + 29.6% +/-7% C. Asian Sarmatian related with a p-value of 0.40 (passing is 0.05)