PDA

View Full Version : MDLP K23b Project - Part 2 of 2- Community PCAs, graphs, & charts



Kurd
03-12-2015, 05:35 PM
This is part 2 of 2 of the MDLP k23b analysis project. The majority of the analysis will be done under this thread. The goals of part 2 of the project will be as follows:

1- To perform analysis via a PCA plot of the community member data. Component 1 of the PCA (x-axis) will consist of the sum of all the S. Asian, SE Asian, and NE Asian MDLP components minus all the West Asian components. Component 2 of the PCA (y-axis) will consist of the sum of all the MDLP European components minus all the African components. This choice of the components will nicely break up populations into clusters somewhat consistent with geographical positioning;

2- Create sorted admixture cluster graphs for the community membership;

3- Create sorted admixture overlap charts for a couple of the major ethnic groups.

I encourage members from all ethnicities to post their data here. To make it less time consuming for me, I would appreciate it if you would copy and paste your data as it appears next to the pie chart graphic, as that is the order I will enter data, and I can thus copy and paste it into my spreadsheet, which will minimize errors. I encourage members to post their data by Saturday midnight US Pacific Time, as I am not able to make more than one update to the charts due to time constraints.

It would be nice if someone can post MDLP K23 admixtures and oracles for some of the newly sequenced ancient genomes, Yamnaya, etc., so that I can incorporate those into the PCAs and graphs.

The following definitions will apply throughout the project:

1- S. Asia = SC Asia + S. India
2- NE & SE Asia = The sum of Amerindian, Ancestral Altaic, Arctic, Tungus Altaic, Paleo Siberian, SE Asian, Australoid, Austronesian, Melano Polynesian
3- W. Asia = Near East + Caucasian
4- The rest is self explanatory

To start here is my data:

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20ADMIX_zpsididmios.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20ADMIX_zpsididmios.jpg.html)

I have highlighted my W. Asian surrogates in green, and S/C Asian ones in brown

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20k23b%20Oracle%20table_zpshi1bbtjn.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20k23b%20Oracle%20table_zpshi1bbtjn.jpg.html)

Kurd
03-12-2015, 05:38 PM
This is my data consolidated into major regions. I will be using these components when I do community cluster graphs.


http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20k23%20Admixture%20CHART_zpsloqmee6w.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20k23%20Admixture%20CHART_zpsloqmee6w.jpg.htm l)

Humanist
03-12-2015, 06:01 PM
My results:


0.64 Amerindian
1.39 Ancestral_Altaic
0.23 Archaic_Human
42.58 Caucasian
8.49 European_Early_Farmers
0.29 European_Hunters_Gatherers
0.64 Melano_Polynesian
18.91 Near_East
4.55 North_African
0.55 Paleo_Siberian
19.68 South_Central_Asian
1.64 South_Indian
0.41 Tungus-Altaic

IQ Mandaean 1


0.94 Amerindian
0.03 Archaic_Human
42.5 Caucasian
8.82 European_Early_Farmers
0.73 European_Hunters_Gatherers
0.57 Melano_Polynesian
17.38 Near_East
5.96 North_African
0.82 Paleo_Siberian
19.71 South_Central_Asian
0.02 South_East_Asian
2.51 South_Indian

IQ Mandaean 2


0.65 Amerindian
1.42 Ancestral_Altaic
0.2 Archaic_Human
0.39 Arctic
0.19 Austronesian
40.73 Caucasian
0.22 East_African
4.53 European_Early_Farmers
20.56 Near_East
8.27 North_African
20.91 South_Central_Asian
1.84 South_Indian

Táltos
03-12-2015, 06:02 PM
Here are my results.
1 European_Hunters_Gatherers 38.55
2 Caucasian 28.32
3 European_Early_Farmers 18.97
4 South_Central_Asian 5.27
5 Near_East 3.19
6 North_African 2.07
7 Ancestral_Altaic 2.06
8 Melano_Polynesian 1.00

4031

4032

Afshar
03-12-2015, 06:03 PM
Population
Amerindian 1.80%
Ancestral_Altaic 3.70%
South_Central_Asian 15.64%
Arctic -
South_Indian 2.81%
Australoid 0.21%
Austronesian 1.04%
Caucasian 40.19%
Archaic_Human 0.06%
East_African -
East_Siberian 0.87%
European_Early_Farmers 8.27%
Khoisan -
Melano_Polynesian -
Archaic_African -
Near_East 12.68%
North_African 1.91%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic 4.85%
European_Hunters_Gatherers 5.97%

Humanist
03-12-2015, 06:19 PM
To make it less time consuming for me, I would appreciate it if you would copy and paste your data as it appears next to the pie chart graphic, as it is alphabetisized, and I can thus copy and paste it into my spreadsheet, which will minimize errors.

Sorry about that!

Dr_McNinja
03-12-2015, 06:24 PM
Population
Amerindian 0.70%
Ancestral_Altaic 4.21%
South_Central_Asian 33.13%
Arctic -
South_Indian 38.30%
Australoid 0.22%
Austronesian -
Caucasian 7.32%
Archaic_Human -
East_African -
East_Siberian 0.83%
European_Early_Farmers 3.55%
Khoisan -
Melano_Polynesian -
Archaic_African 0.09%
Near_East 1.54%
North_African -
Paleo_Siberian 2.45%
African_Pygmy -
South_East_Asian -
Subsaharian 0.43%
Tungus-Altaic 0.91%
European_Hunters_Gatherers 6.31%

MfA
03-12-2015, 06:29 PM
.....

NK19191
03-12-2015, 06:37 PM
Population
Amerindian -
Ancestral_Altaic 3.45%
South_Central_Asian 25.24%
Arctic 1.11%
South_Indian 6.00%
Australoid 0.16%
Austronesian -
Caucasian 35.50%
Archaic_Human 0.39%
East_African 0.11%
East_Siberian 0.07%
European_Early_Farmers 5.92%
Khoisan -
Melano_Polynesian 0.58%
Archaic_African -
Near_East 12.41%
North_African 1.23%
Paleo_Siberian -
African_Pygmy 0.32%
South_East_Asian 2.06%
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 5.46%

Dr_McNinja
03-12-2015, 06:46 PM
Ancestral Altaic is actually leftover ANE, so that's not really East Asian. You could join it with European but not sure that would be a lot better since EEF + WHG together is pretty much middle neolithic Gok2-like European admixture. With ANE it could approximate something else, or just treat it separately altogether like "North Eurasian" since it seems to vary a lot (the other components themselves use up most of the ANE).

It depends if the European users are getting a lot of it which they don't seem to, so it just sticks out like a sore thumb. In South Asians some of that would've gone into European in other calculators, so it shows there's a discrepancy between what's actually European-like and what Admixture lumps into European because of ANE in other calculators for South Asians.

Or I guess you could put it into South Asian though since it appears that for everyone outside South Asia, it may be coming from South Asian components, but for everyone inside South Asia, it's coming from Gedrosian or European.

But treating it separately as 'North Eurasian' would allow for useful comparisons and avoid committing to one or the other category where it doesn't seem to fit into any.

Rukha
03-12-2015, 07:22 PM
Population
Amerindian 2.58%
Ancestral_Altaic 8.47%
South_Central_Asian 34.15%
Arctic -
South_Indian 16.72%
Australoid 0.61%
Austronesian -
Caucasian 21.50%
Archaic_Human -
East_African -
East_Siberian -
European_Early_Farmers -
Khoisan -
Melano_Polynesian -
Archaic_African -
Near_East 2.13%
North_African 0.21%
Paleo_Siberian -
African_Pygmy -
South_East_Asian 0.87%
Subsaharian 0.62%
Tungus-Altaic 1.84%
European_Hunters_Gatherers 10.30%



Least-squares method.

Using 1 population approximation:
1 Afghan_Pushtun @ 6.022763
2 Pashtun_Afghani @ 6.598314
3 Tajik_Pomiri_Ishkashim @ 7.877693
4 Pakistani_Pushtun @ 10.038360
5 Tajik_Pomiri_Shugnan @ 10.918867
6 Tajik_Pomiri_Rushan @ 11.797170
7 Uzbek_Afghan @ 13.621663
8 Tajik_Afghan @ 14.289532
9 Pathan @ 15.131264
10 Tajik_Yagnobi @ 15.695024
11 Parsi @ 16.078205
12 Burusho @ 18.395235
13 Jatt_Haryana @ 19.676146
14 Jatt_Pahari @ 20.610188
15 Iranian @ 22.572412
16 Punjabi_Gujjar @ 22.696201
17 Sindhi @ 24.821438
18 Jatt_Muslim @ 25.668371
19 Turkmen_Uzbekistan @ 25.870001
20 Mumbai_Jew @ 25.994358

Using 2 populations approximation:
1 50% Pakistani_Pushtun +50% Tajik_Pomiri_Shugnan @ 4.334452

Using 3 populations approximation:
1 50% Pathan +25% Tabassaran +25% Tajik_Pomiri_Ishkashim @ 3.331667

Kurd
03-12-2015, 07:25 PM
Or I guess you could put it into South Asian though since it appears that for everyone outside South Asia, it may be coming from South Asian components

For the PCA, it would not make a difference in that regard for most members, since the 1st half of component 1 is (S/SC Asian + E. Asian). For cluster graphs, I could label one component N/NE/SE Asian, and it would thus get included in there along with Siberian.


EDIT: As an alternative to the above, I may leave out Ancestral Altaic from the PCA altogether

Stephen1986
03-12-2015, 07:55 PM
I didn't include this in my post in the previous thread, so here's the alphabetical list -

Amerindian 0.46%
Ancestral_Altaic 5.38%
South_Central_Asian 6.05%
Arctic -
South_Indian 0.07%
Australoid -
Austronesian -
Caucasian 20.52%
Archaic_Human -
East_African -
East_Siberian -
European_Early_Farmers 29.43%
Khoisan -
Melano_Polynesian 0.49%
Archaic_African -
Near_East -
North_African 0.97%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 36.57%

fil
03-12-2015, 08:36 PM
Amerindian 0.17%
Ancestral_Altaic 1.78%
South_Central_Asian 31.73%
Arctic 0.13%
South_Indian 48.89%
Australoid 1.57%
Austronesian 0.60%
Caucasian 8.18%
Archaic_Human -
East_African 0.88%
East_Siberian 0.15%
European_Early_Farmers -
Khoisan -
Melano_Polynesian 0.79%
Archaic_African -
Near_East 2.15%
North_African 0.67%
Paleo_Siberian -
African_Pygmy -
South_East_Asian 0.88%
Subsaharian -
Tungus-Altaic 0.88%
European_Hunters_Gatherers 0.53%

ViktorL1
03-12-2015, 09:12 PM
Population
Amerindian 0.92%
Ancestral_Altaic 6.45%
South_Central_Asian 17.99%
Arctic 1.50%
South_Indian 7.17%
Australoid -
Austronesian -
Caucasian 17.94%
Archaic_Human 0.44%
East_African -
East_Siberian 6.80%
European_Early_Farmers -
Khoisan -
Melano_Polynesian -
Archaic_African 0.23%
Near_East 0.66%
North_African 0.23%
Paleo_Siberian 1.27%
African_Pygmy -
South_East_Asian 8.78%
Subsaharian -
Tungus-Altaic 17.88%
European_Hunters_Gatherers 11.73%


Using 1 population approximation:
1 Tajik_Tajikistan @ 1.482572
2 Uzbek @ 4.865723
3 Turkmen_Afghan @ 8.899322
4 Uygur-Han @ 10.646794
5 Turkmen_Uzbekistan @ 11.425840
6 Uygur @ 12.460958
7 Hazara @ 12.748605
8 Karakalpak @ 13.170482
9 Tajik_Afghan @ 16.408781
10 Tatar-Siberian @ 18.339470
11 Uzbek_Afghan @ 18.357813
12 Kazahs @ 18.734402
13 Crimean_Tatar_Step @ 20.293940
14 Nogai @ 20.969976
15 Kazakh @ 22.069340
16 Bashkir @ 24.320910
17 Tajik_Pomiri_Rushan @ 25.116444
18 Kyrgyz_Bishkek @ 25.353327
19 Tajik_Pomiri_Shugnan @ 26.363405
20 Tajik_Pomiri_Ishkashim @ 26.530727




Using 4 populations approximation:
1 Tajik_Tajikistan + Tajik_Tajikistan + Tajik_Tajikistan + Uzbek @ 1.321406
2 Tajik_Tajikistan + Tajik_Tajikistan + Tajik_Tajikistan + Tajik_Tajikistan @ 1.482572
3 Tajik_Tajikistan + Tajik_Tajikistan + Uzbek + Uzbek @ 2.275851
4 Tajik_Tajikistan + Tajik_Tajikistan + Tajik_Tajikistan + Uygur-Han @ 2.355236
5 Tajik_Tajikistan + Tajik_Tajikistan + Tajik_Tajikistan + Turkmen_Afghan @ 2.435131
6 Tajik_Tajikistan + Tajik_Tajikistan + Tajik_Tajikistan + Turkmen_Uzbekistan @ 2.720558
7 Karakalpak + Tajik_Afghan + Tajik_Tajikistan + Tajik_Tajikistan @ 2.758465
8 Karakalpak + Tajik_Tajikistan + Tajik_Tajikistan + Uzbek_Afghan @ 3.024047
9 Tajik_Tajikistan + Tajik_Tajikistan + Turkmen_Afghan + Uygur-Han @ 3.046044
10 Burusho + Crimean_Tatar_Step + Lak + Mongol_Khalkha @ 3.111164
11 Tajik_Tajikistan + Tajik_Tajikistan + Turkmen_Afghan + Uzbek @ 3.118806
12 Tajik_Tajikistan + Tajik_Tajikistan + Tajik_Tajikistan + Uygur @ 3.144663
13 Tajik_Tajikistan + Tajik_Tajikistan + Uygur-Han + Uzbek @ 3.166412
14 Burusho + Crimean_Tatar_Step + Dargin_Urkarah + Mongol_Khalkha @ 3.182810
15 Avar + Burusho + Crimean_Tatar_Step + Mongol_Khalkha @ 3.191904
16 Crimean_Tatar_Step + Hazara + Tajik_Afghan + Tajik_Tajikistan @ 3.236356
17 Burusho + Crimean_Tatar_Step + Mongol_Khalkha + Tabassaran @ 3.244256
18 Afghan_Pushtun + Karakalpak + Karakalpak + Tajik_Tajikistan @ 3.296244
19 Hazara + Tajik_Tajikistan + Tajik_Tajikistan + Tajik_Tajikistan @ 3.306517
20 Kazakh + Tajik_Pomiri_Rushan + Tajik_Tajikistan + Tajik_Tajikistan @ 3.312823

Hope this helps.

randwulf
03-12-2015, 10:43 PM
I wasn't clear if those of us that had already posted the results prior should re-post. Just in case, here it is:

Population
Amerindian 0.36%
Ancestral_Altaic 2.37%
South_Central_Asian 4.40%
Arctic 0.09%
South_Indian -
Australoid -
Austronesian -
Caucasian 23.13%
Archaic_Human -
East_African -
East_Siberian 0.93%
European_Early_Farmers 30.27%
Khoisan -
Melano_Polynesian 0.84%
Archaic_African 0.28%
Near_East 2.66%
North_African 1.36%
Paleo_Siberian -
African_Pygmy 0.09%
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 33.22%

Gray Fox
03-12-2015, 11:09 PM
Here you go..

American of primarily British background. Next largest ethnicity is southern German followed by smaller amounts of French, Welsh and Irish. Small potential (less than one percent) for Native American or West African admix. I can't determine if these less than one percent regions are real or noise.

Population
Amerindian -
Ancestral_Altaic 3.61%
South_Central_Asian 7.61%
Arctic -
South_Indian -
Australoid 0.99%
Austronesian -
Caucasian 20.03%
Archaic_Human -
East_African -
East_Siberian -
European_Early_Farmers 27.81%
Khoisan -
Melano_Polynesian -
Archaic_African 0.93%
Near_East 1.16%
North_African 0.60%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 37.24%

Rukha
03-13-2015, 12:08 AM
Kit M020637 (Yamnaya_Sok_River I0443)


Population
Amerindian 1.97%
Ancestral_Altaic 14.24%
South_Central_Asian 23.46%
Arctic 0.61%
South_Indian -
Australoid -
Austronesian -
Caucasian 17.27%
Archaic_Human 0.13%
East_African -
East_Siberian -
European_Early_Farmers -
Khoisan 0.19%
Melano_Polynesian -
Archaic_African -
Near_East -
North_African -
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 42.13%


Least-squares method.

Using 1 population approximation:
1 Tatar_Mishar @ 21.040140
2 Tatar @ 23.675005
3 Tatar-Mishar @ 23.889797
4 Moksha @ 23.984941
5 Tatar-Kazan @ 24.030188
6 Komi @ 24.391094
7 Tatar_Kryashen @ 24.504982
8 Erzya @ 24.563124
9 Russian-Ural @ 24.890877
10 Chuvashs @ 25.529049
11 Russian_Meshtchyora @ 26.332289
12 Chuvash @ 26.863770
13 Finn_East @ 26.953863
14 Russian-Upper-Volga @ 27.135542
15 Swede_Saami @ 27.159199
16 Mordovian @ 27.363317
17 Finn_West @ 27.429829
18 Finnish-East @ 27.589832
19 Tatar-Lithuanian @ 27.646143
20 Russian-North @ 27.892235

Using 2 populations approximation:
1 50% Karelian +50% Tajik_Pomiri_Shugnan @ 11.530635


Using 3 populations approximation:
1 50% Estonian +25% Kalash +25% Udmurd @ 10.216213

vettor
03-13-2015, 12:24 AM
here is mine

Population
Amerindian -
Ancestral_Altaic 3.04%
South_Central_Asian 5.25%
Arctic -
South_Indian -
Australoid -
Austronesian -
Caucasian 28.43%
Archaic_Human -
East_African -
East_Siberian -
European_Early_Farmers 28.49%
Khoisan -
Melano_Polynesian 0.98%
Archaic_African -
Near_East 4.36%
North_African 4.33%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 25.13%

WOW......25% EHG............I thought everyone said that T ydna where only farmers!!!!


wow again ...............big Generic Distance difference between #1 and #2

Using 1 population approximation:
1 Italian_North @ 0.960944
2 Kosovar @ 10.069535
3 Greek_Northwest @ 10.398839
4 Italian_Piedmont @ 10.514088
5 German-Volga @ 11.196369
6 Italian_Tuscan @ 11.257127
7 Italian_Bergamo @ 11.277226
8 Bulgarian @ 12.858762
9 Greek_Thessaly @ 12.882842
10 Albanian_Tirana @ 12.891836
11 French @ 13.053646
12 Greek_Thessaloniki @ 13.376898
13 Greek_Peloponnesos @ 13.556684
14 South_German @ 14.155749
15 Macedonian @ 14.420380
16 Montenegrian @ 14.452551
17 Serb_Serbia @ 14.741122
18 Belgian @ 14.777686
19 Spanish_Baleares_IBS @ 14.992661
20 English @ 15.021181

Using 2 populations approximation:
1 50% Italian_North +50% Italian_North @ 0.960944

Kurd
03-13-2015, 12:49 AM
KARELIAN HG:

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20karelian_zpssmtofa6b.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20karelian_zpssmtofa6b.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20karelian%20Oracle_zpswhhqdvws.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20karelian%20Oracle_zpswhhqdvws.jpg.htm l)

EDIT: If someone has time, I would appreciate it if some other ancients, such as Ust-Ichim, etc could get posted. I guess we should also check accuracy of the results by checking number of SNPs evaluated by the calculator.

Agamemnon
03-13-2015, 01:03 AM
My mother:

Population
Amerindian -
Ancestral_Altaic 4.27%
South_Central_Asian 5.10%
Arctic -
South_Indian 0.17%
Australoid -
Austronesian -
Caucasian 22.72%
Archaic_Human 0.13%
East_African -
East_Siberian -
European_Early_Farmers 26.94%
Khoisan -
Melano_Polynesian -
Archaic_African -
Near_East 0.85%
North_African 1.68%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 38.10%

Using 1 population approximation:
1 Belgian @ 2.285868
2 Frisian @ 2.974171
3 South_German @ 2.981146
4 Dutch @ 3.063809
5 English @ 3.336424
6 North_European @ 4.040587
7 Irish @ 4.093512
8 German-Volga @ 4.863644
9 English_Kent_GBR @ 5.674329
10 CEU @ 6.067739
11 Norwegian_East @ 6.123525
12 North_German @ 6.460650
13 British @ 6.828248
14 English_Cornwall_GBR @ 6.847414
15 Dane @ 6.913705
16 Welsh @ 7.195769
17 Norwegian_West @ 7.345781
18 Icelandic @ 7.731325
19 Scottish_Argyll_Bute_GBR @ 7.815835
20 Swede @ 7.938396

Using 2 populations approximation:
1 50% Frisian +50% South_German @ 1.333866


Using 3 populations approximation:
1 50% Dutch +25% Kashub +25% Spanish_Cataluna_IBS @ 0.785721


Myself:

Population
Amerindian -
Ancestral_Altaic 2.90%
South_Central_Asian 6.68%
Arctic -
South_Indian 0.29%
Australoid 0.34%
Austronesian 0.50%
Caucasian 29.36%
Archaic_Human 0.28%
East_African -
East_Siberian -
European_Early_Farmers 23.05%
Khoisan -
Melano_Polynesian -
Archaic_African 0.14%
Near_East 5.89%
North_African 3.92%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 26.62%

Using 1 population approximation:
1 Italian_North @ 5.576355
2 Bulgarian @ 8.718830
3 Greek_Northwest @ 9.484306
4 Kosovar @ 9.964145
5 Macedonian @ 10.354059
6 German-Volga @ 10.494763
7 Montenegrian @ 10.652430
8 Serb_Serbia @ 11.466665
9 Greek_Thessaloniki @ 11.521767
10 Greek_Thessaly @ 11.722589
11 Greek_Peloponnesos @ 11.877713
12 Albanian_Tirana @ 12.008486
13 Romanian @ 12.269452
14 Gagauz @ 13.108708
15 South_German @ 13.666821
16 Central_Greek @ 14.663303
17 Italian_Tuscan @ 14.720754
18 German_East @ 14.806586
19 Italian_Piedmont @ 14.922957
20 Austrian @ 15.335489

Using 2 populations approximation:
1 50% Cretan +50% Norwegian_East @ 1.088383


Using 3 populations approximation:
1 50% Cretan +25% Icelandic +25% Norwegian_East @ 0.990803

Kurd
03-13-2015, 04:00 AM
Yamna IO231:

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20YAMNA%20IO231_zpsjoypkxu3.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20YAMNA%20IO231_zpsjoypkxu3.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20YAMNA%20IO231%20ORACLE_zpsqxuxrf3e.jp g (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20YAMNA%20IO231%20ORACLE_zpsqxuxrf3e.jp g.html)

Kurd
03-13-2015, 04:04 AM
UST-ISHIM, SIBERIA, 45KY:

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20k23%20Ust%20Ishim%20Siberia_zps49zq46rv.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20k23%20Ust%20Ishim%20Siberia_zps49zq46rv.jpg .html)

Admix Results (sorted):
# Population Percent
1 South_Indian 28.37
2 Australoid 8.74
3 South_East_Asian 7.00
4 European_Early_Farmers 6.64
5 Austronesian 6.54
6 European_Hunters_Gatherers 5.94
7 East_African 4.83
8 South_Central_Asian 4.78
9 Melano_Polynesian 4.61
10 Archaic_African 4.14
11 Near_East 3.02
12 Subsaharian 2.93
13 North_African 2.57
14 Paleo_Siberian 2.34
15 Tungus-Altaic 2.04
16 Arctic 1.44
17 Ancestral_Altaic 1.28
18 Archaic_Human 1.12
19 Amerindian 1.11


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Ayta_AE @ 24.783888
2 Nepalese @ 27.270012
3 Kensiu @ 29.925926
4 Tamil_Singapore @ 29.993927
5 Cochin_Jew @ 30.449663
6 Onge @ 31.218340
7 Pahari @ 31.865376
8 Mumbai_Jew @ 33.207047
9 Dhaka_mixed_popul @ 33.615238
10 Hindi @ 34.281342
11 Uygur-Han @ 34.523560
12 Jatt_Haryana @ 34.819286
13 Tiwari @ 35.455795
14 Pakistani_Pushtun @ 35.725113
15 Marathi @ 35.728867
16 Jatt_Muslim @ 35.853012
17 Mamanawa @ 36.317513
18 Vaish @ 36.851608
19 Brahmins_UP @ 36.884640
20 Spiti @ 37.032146

Using 2 populations approximation:
1 50% Onge +50% Puerto_Rican @ 13.927173

PureEvil
03-13-2015, 05:05 AM
My results:
# Population Percent
1 European_Hunters_Gatherers 36.3
2 European_Early_Farmers 26.7
3 Caucasian 22.68
4 Ancestral_Altaic 5.66
5 South_Central_Asian 4.82
6 North_African 1.97
7 South_Indian 1.16
8 Arctic 0.34
9 Archaic_Human 0.2
10 Khoisan 0.11
11 Amerindian 0.07

Silesian
03-13-2015, 06:30 AM
My results-Poland
# Population Percent
1 European_Hunters_Gatherers 48.61
2 Caucasian 27.23
3 European_Early_Farmers 13.91
4 South_Central_Asian 3.87
5 Ancestral_Altaic 2.91
6 Paleo_Siberian 0.74
7 South_Indian 0.66
8 East_Siberian 0.64
9 Near_East 0.5
10 Austronesian 0.43
11 Melano_Polynesian 0.33
12 Australoid 0.15
13 Archaic_Human 0.03

Samara Hunter Gatherer-Russia-Io443-M020637

# Population Percent
1 European_Hunters_Gatherers 42.13
2 South_Central_Asian 23.46
3 Caucasian 17.27
4 Ancestral_Altaic 14.24
5 Amerindian 1.97
6 Arctic 0.61
7 Khoisan 0.19
8 Archaic_Human 0.13

Russian sample 1



# Population Percent
1 European_Hunters_Gatherers 49.37
2 Caucasian 29.41
3 European_Early_Farmers 11.39
4 South_Central_Asian 3.62
5 Ancestral_Altaic 2.66
6 Tungus-Altaic 1.2
7 Austronesian 1.11
8 Arctic 0.47
9 East_Siberian 0.34
10 Melano_Polynesian 0.24
11 Archaic_Human 0.13
12 South_Indian 0.08


Russian sample 2


# Population Percent
1 European_Hunters_Gatherers 52.02
2 Caucasian 17.17
3 European_Early_Farmers 13.29
4 Ancestral_Altaic 6.57
5 East_Siberian 3.69
6 Arctic 1.66
7 Austronesian 1.48
8 South_Central_Asian 1.46
9 Melano_Polynesian 1.09
10 Amerindian 0.66
11 Khoisan 0.41
12 North_African 0.39
13 Archaic_Human 0.12

Baltic sample

# Population Percent
1 European_Hunters_Gatherers 54.24
2 Caucasian 29.36
3 European_Early_Farmers 9.02
4 Ancestral_Altaic 1.99
5 Near_East 1.05
6 South_Indian 0.95
7 East_Siberian 0.58
8 Austronesian 0.54
9 Amerindian 0.46
10 Tungus-Altaic 0.43
11 South_Central_Asian 0.29
12 South_East_Asian 0.27
13 Subsaharian 0.26
14 Australoid 0.26
15 Arctic 0.18
16 Archaic_African 0.12

DMXX
03-13-2015, 06:59 AM
My full results.



# Population Percent
1 Caucasian 34.81
2 South_Central_Asian 23.74
3 Near_East 12.05
4 European_Hunters_Gatherers 6.48
5 European_Early_Farmers 5.7
6 South_Indian 4
7 Ancestral_Altaic 2.68
8 Tungus-Altaic 2.45
9 North_African 1.78
10 South_East_Asian 1.61
11 East_African 1.43
12 Paleo_Siberian 0.81
13 Australoid 0.72
14 Arctic 0.64
15 Amerindian 0.4
16 Subsaharian 0.39
17 Archaic_African 0.28
18 Archaic_Human 0.04


The requested breakdown (also included the NE&SE Asian score minus Ancestral Altaic as discussed on page 2 for your convenience):



S. Asia 27.74
NE&SE Asia 9.31
NE&SE Asia(-Anc.Alt.) 6.63
W. Asia 46.86
Euro&Med 12.18
African 3.92

Salkin
03-13-2015, 10:57 AM
Amerindian 0.55%
Ancestral_Altaic 5.17%
South_Central_Asian 2.56%
Arctic 0.42%
South_Indian 0.55%
Australoid -
Austronesian 0.09%
Caucasian 19.90%
Archaic_Human -
East_African -
East_Siberian -
European_Early_Farmers 24.64%
Khoisan -
Melano_Polynesian 0.36%
Archaic_African -
Near_East -
North_African 1.78%
Paleo_Siberian 0.35%
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 43.63%

Kurd
03-13-2015, 12:09 PM
IRANIAN BALOCH 2:

Population
Amerindian 2.15%
Ancestral_Altaic 4.04%
South_Central_Asian 37.74%
Arctic -
South_Indian 12.88%
Australoid -
Austronesian -
Caucasian 22.29%
Archaic_Human 0.07%
East_African 1.13%
East_Siberian -
European_Early_Farmers -
Khoisan -
Melano_Polynesian 0.49%
Archaic_African 0.46%
Near_East 11.64%
North_African 2.24%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian 2.58%
Tungus-Altaic -
European_Hunters_Gatherers 2.29%


Admix Results (sorted):
# Population Percent
1 South_Central_Asian 37.74
2 Caucasian 22.29
3 South_Indian 12.88
4 Near_East 11.64
5 Ancestral_Altaic 4.04
6 Subsaharian 2.58
7 European_Hunters_Gatherers 2.29
8 North_African 2.24
9 Amerindian 2.15
10 East_African 1.13


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Parsi @ 10.414766
2 Afghan_Pushtun @ 12.683915
3 Pashtun_Afghani @ 13.153192
4 Iranian @ 15.518795
5 Uzbek_Afghan @ 16.949854
6 Pakistani_Pushtun @ 17.094519
7 Tajik_Pomiri_Ishkashim @ 17.717024
8 Tajik_Afghan @ 18.932661
9 Pathan @ 19.762735
10 Tajik_Pomiri_Shugnan @ 20.300516
11 Tajik_Pomiri_Rushan @ 20.331760
12 Tajik_Yagnobi @ 21.135485
13 Kurd_South @ 21.838951
14 Kurd_East @ 21.947428
15 Makrani @ 22.079990
16 Burusho @ 23.428642
17 Kurd @ 23.887026
18 Balochi @ 25.275084
19 Jatt_Pahari @ 25.491989
20 Kurd_North @ 25.512718

Using 2 populations approximation:
1 50% Kurd +50% Sindhi @ 7.277781


Using 3 populations approximation:
1 50% Iranian +25% Makrani +25% Pathan @ 4.792076

Kurd
03-13-2015, 12:11 PM
IRANIAN QESHIM ISLAND:

QESHMI
Population
Amerindian -
Ancestral_Altaic 3.17%
South_Central_Asian 36.23%
Arctic 0.61%
South_Indian 12.10%
Australoid -
Austronesian -
Caucasian 23.39%
Archaic_Human 0.42%
East_African 0.92%
East_Siberian 1.59%
European_Early_Farmers 2.57%
Khoisan -
Melano_Polynesian 0.31%
Archaic_African -
Near_East 12.13%
North_African 0.71%
Paleo_Siberian -
African_Pygmy 0.38%
South_East_Asian -
Subsaharian 1.65%
Tungus-Altaic -
European_Hunters_Gatherers 3.82%


Admix Results (sorted):
# Population Percent
1 South_Central_Asian 36.23
2 Caucasian 23.39
3 Near_East 12.13
4 South_Indian 12.10
5 European_Hunters_Gatherers 3.82
6 Ancestral_Altaic 3.17
7 European_Early_Farmers 2.57
8 Subsaharian 1.65
9 East_Siberian 1.59


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Parsi @ 8.979746
2 Afghan_Pushtun @ 12.132631
3 Pashtun_Afghani @ 13.279145
4 Iranian @ 13.366030
5 Uzbek_Afghan @ 16.024700
6 Pakistani_Pushtun @ 17.331367
7 Tajik_Pomiri_Ishkashim @ 17.534542
8 Tajik_Afghan @ 17.884474
9 Tajik_Pomiri_Rushan @ 19.169098
10 Kurd_South @ 19.226887
11 Kurd_East @ 19.475214
12 Tajik_Pomiri_Shugnan @ 19.696789
13 Tajik_Yagnobi @ 19.744432
14 Pathan @ 20.984758
15 Kurd @ 21.631931
16 Kurd_North @ 22.914423
17 Azeri @ 23.505789
18 Baku_WGA @ 23.610964
19 Makrani @ 23.979609
20 Ain_Touta_WGA @ 24.444349

Using 2 populations approximation:
1 50% Iranian +50% Pashtun_Afghani @ 6.377221


Using 3 populations approximation:
1 50% Iranian +25% Makrani +25% Pakistani_Pushtun @ 4.208208

parasar
03-13-2015, 02:47 PM
My results:

Population
Amerindian 1.23%
Ancestral_Altaic 3.54%
South_Central_Asian 29.99%
Arctic 1.43%
South_Indian 42.23%
Australoid 1.46%
Austronesian 0.61%
Caucasian 8.71%
Archaic_Human 0.06%
East_African -
East_Siberian 0.05%
European_Early_Farmers -
Khoisan -
Melano_Polynesian 0.61%
Archaic_African -
Near_East 1.30%
North_African 0.53%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 8.26%

Hanna
03-13-2015, 03:14 PM
My scores:

Population
Amerindian 0.32%
Ancestral_Altaic 2.82%
South_Central_Asian 18.16%
Arctic 0.41%
South_Indian 0.13%
Australoid 0.29%
Austronesian -
Caucasian 45.98%
Archaic_Human -
East_African -
East_Siberian 2.20%
European_Early_Farmers 11.37%
Khoisan 0.10%
Melano_Polynesian 0.60%
Archaic_African 0.30%
Near_East 9.52%
North_African 0.77%
Paleo_Siberian -
African_Pygmy -
South_East_Asian 0.58%
Subsaharian 0.51%
Tungus-Altaic 2.81%
European_Hunters_Gatherers 3.14%


Oracle 4:

Admix Results (sorted):

# Population Percent
1 Caucasian 45.98
2 South_Central_Asian 18.16
3 European_Early_Farmers 11.37
4 Near_East 9.52
5 European_Hunters_Gatherers 3.14
6 Ancestral_Altaic 2.82
7 Tungus-Altaic 2.81
8 East_Siberian 2.20


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Turk_Kayseri @ 4.266644
2 Turk @ 6.815643
3 Turk_Adana @ 8.286249
4 Armenian @ 8.361481
5 Georgian_Jew @ 8.882082
6 Armenian_Yerevan @ 9.504634
7 Assyrian_Iraqi @ 9.563898
8 Jew_Tat @ 10.325316
9 Uzbek_Tashkent @ 10.710565
10 Uzbekistani_Jew @ 10.758110
11 Turk_Istanbul @ 10.797637
12 Azeri @ 10.807470
13 Assyrian_Arzni @ 10.868814
14 Kurd_North @ 11.274456
15 Iraqi_Chaldean @ 11.923533
16 Cirkassian @ 12.250082
17 Turk_Aydin @ 12.944167
18 Kurd_Jew @ 13.072672
19 Baku_WGA @ 13.332445
20 Iranian_Jew @ 13.429023

Using 2 populations approximation:
1 50% Armenian +50% Turk_Kayseri @ 3.999555


Using 3 populations approximation:
1 50% Azeri +25% Greek_Smyrna +25% Turk_Trabzon @ 3.575541

Hanna
03-13-2015, 03:27 PM
My aunt's scores:

Population
Amerindian 0.64%
Ancestral_Altaic 2.06%
South_Central_Asian 21.87%
Arctic -
South_Indian -
Australoid -
Austronesian -
Caucasian 47.43%
Archaic_Human -
East_African -
East_Siberian 1.33%
European_Early_Farmers 9.56%
Khoisan -
Melano_Polynesian 1.52%
Archaic_African -
Near_East 10.59%
North_African 3.03%
Paleo_Siberian -
African_Pygmy -
South_East_Asian 0.15%
Subsaharian -
Tungus-Altaic 0.21%
European_Hunters_Gatherers 1.61%


Oracle 4:

Admix Results (sorted):

# Population Percent
1 Caucasian 47.43
2 South_Central_Asian 21.87
3 Near_East 10.59
4 European_Early_Farmers 9.56
5 North_African 3.03
6 Ancestral_Altaic 2.06
7 European_Hunters_Gatherers 1.61
8 Melano_Polynesian 1.52
9 East_Siberian 1.33


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Armenian @ 6.630957
2 Armenian_Yerevan @ 7.356857
3 Turk_Kayseri @ 8.268700
4 Uzbekistani_Jew @ 8.346710
5 Assyrian_Arzni @ 8.502128
6 Jew_Tat @ 8.530933
7 Assyrian_Iraqi @ 8.894340
8 Georgian_Jew @ 9.137982
9 Uzbek_Tashkent @ 9.448464
10 Kurd_North @ 9.573703
11 Azeri @ 9.615912
12 Turk_Adana @ 10.245461
13 Kurd_Jew @ 10.808213
14 Turk @ 10.862277
15 Iraqi_Chaldean @ 11.099289
16 Baku_WGA @ 11.105381
17 Kurd @ 11.106915
18 Iranian_Jew @ 11.182240
19 Iraqi_Mandean @ 12.135379
20 Kurd_East @ 12.458428

Using 2 populations approximation:
1 50% Armenian +50% Azeri @ 4.121730


Using 3 populations approximation:
1 50% Armenian +25% Kurd +25% Turk_Kayseri @ 3.418141

Kurd
03-13-2015, 03:33 PM
Zarafshaan, Iranian Baloch's results. His results are similar to mine, except he has higher SC Asian (actually may be very close to the MDLP SC Asian reference sample), and I have higher E. Eurasian components. I also can't help but notice that PARSI seems to show up as one of the best surrogates for me, him, and I believe Jesus. This is quiet interesting in lieu of the fact that the Parsis have been in India for centuries. This may be evidence of the Parsis low rate of mixing with the local Indian population.

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20ZARAFSHAN_zpseemd3hsz.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20ZARAFSHAN_zpseemd3hsz.jpg.html)

Admix Results (sorted):
# Population Percent
1 South_Central_Asian 37.10
2 Caucasian 23.75
3 Near_East 11.82
4 South_Indian 10.69
5 Ancestral_Altaic 4.80
6 North_African 2.86
7 Subsaharian 2.68
8 European_Hunters_Gatherers 2.19
9 Amerindian 1.66


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Parsi @ 10.972406
2 Iranian @ 13.571710
3 Afghan_Pushtun @ 13.649196
4 Pashtun_Afghani @ 14.287457
5 Uzbek_Afghan @ 17.024031
6 Tajik_Pomiri_Ishkashim @ 18.213661
7 Pakistani_Pushtun @ 18.771530
8 Tajik_Afghan @ 18.814621
9 Kurd_South @ 19.682087
10 Tajik_Yagnobi @ 19.748482
11 Kurd_East @ 19.780668
12 Tajik_Pomiri_Rushan @ 19.955227
13 Tajik_Pomiri_Shugnan @ 20.109358
14 Kurd @ 21.547939
15 Pathan @ 22.115313
16 Kurd_North @ 23.378531
17 Baku_WGA @ 23.494114
18 Makrani @ 23.561829
19 Azeri @ 23.678303
20 Ain_Touta_WGA @ 24.472485

Using 2 populations approximation:
1 50% Baku_WGA +50% Makrani @ 6.432648


Using 3 populations approximation:
1 50% Iranian +25% Makrani +25% Pashtun_Afghani @ 5.181563

icebreaker
03-13-2015, 03:41 PM
MDLP K23b Oracle

Admix Results (sorted):

# Population Percent
1 Caucasian 40.38
2 South_Central_Asian 17.22
3 Near_East 11.30
4 European_Early_Farmers 9.45
5 European_Hunters_Gatherers 4.78
6 Tungus-Altaic 4.01
7 South_East_Asian 3.51
8 Ancestral_Altaic 3.28
9 East_Siberian 2.30
10 North_African 1.25
11 South_Indian 1.09


MDLP K23b 4-Ancestors Oracle

Using 1 population approximation:
1 Turk_Kayseri @ 4.945309
2 Turk_Adana @ 5.162985
3 Turk @ 5.646211
4 Georgian_Jew @ 8.315325
5 Azeri @ 8.942010
6 Uzbek_Tashkent @ 9.535335
7 Turk_Istanbul @ 9.611263
8 Turk_Aydin @ 9.826830
9 Uzbekistani_Jew @ 10.267972
10 Turk_Balikesir @ 11.774841
11 Jew_Tat @ 11.988594
12 Kurd_Jew @ 12.271308
13 Baku_WGA @ 12.652429
14 Kurd @ 12.944852
15 Iranian_Jew @ 13.027428
16 Armenian @ 13.489219
17 Armenian_Yerevan @ 13.874441
18 Iraqi_Jew @ 14.061293
19 Cirkassian @ 14.149807
20 Assyrian_Arzni @ 14.278120

Using 2 populations approximation:
1 50% Azeri +50% Turk_Aydin @ 4.093474


Using 3 populations approximation:
1 50% Armenian_Yerevan +25% Greek_Smyrna +25% Uzbek @ 2.979132


Using 4 populations approximation:
1 Greek_Smyrna + Kakheti + Kurd_Jew + Uzbek @ 2.613519
2 Greek_Smyrna + Kakheti + Kurd_Jew + Turkmen_Uzbekistan @ 2.618598
3 Greek_Smyrna + Iranian_Jew + Kakheti + Uzbek @ 2.627265
4 Greek_Smyrna + Iraqi_Jew + Kakheti + Uzbek @ 2.708884
5 Armenian_Yerevan + Greek_Smyrna + Turkmen_Uzbekistan + Uzbekistani_Jew @ 2.760109
6 Greek_Smyrna + Iranian_Jew + Kakheti + Turkmen_Uzbekistan @ 2.765227
7 Armenian_Yerevan + Greek_Smyrna + Turkmen_Uzbekistan + Uzbek_Tashkent @ 2.767456
8 Greek_Smyrna + Kakheti + Turkmen_Uzbekistan + Uzbekistani_Jew @ 2.780003
9 Armenian_Yerevan + Greek_Islands + Turkmen_Uzbekistan + Uzbek_Tashkent @ 2.787662
10 Armenian + Greek_Smyrna + Turkmen_Uzbekistan + Uzbek_Tashkent @ 2.805794
11 Armenian_Yerevan + Baku_WGA + Greek_Islands + Turkmen_Uzbekistan @ 2.823035
12 Armenian + Baku_WGA + Greek_Islands + Turkmen_Uzbekistan @ 2.824483
13 Assyrian_Arzni + Greek_Smyrna + Turkmen_Uzbekistan + Uzbek_Tashkent @ 2.845850
14 Armenian + Georgian_Jew + Turk_Kayseri + Turkmen_Uzbekistan @ 2.854328
15 Assyrian_Arzni + Baku_WGA + Greek_Islands + Turkmen_Uzbekistan @ 2.857517
16 Armenian_Yerevan + Azeri + Greek_Islands + Turkmen_Uzbekistan @ 2.858602
17 Armenian + Cretan + Turkmen_Uzbekistan + Uzbek_Tashkent @ 2.865663
18 Armenian + Greek_Smyrna + Turkmen_Uzbekistan + Uzbekistani_Jew @ 2.867793
19 Armenian + Greek_Islands + Turkmen_Uzbekistan + Uzbek_Tashkent @ 2.876598
20 Armenian_Yerevan + Azeri + Cretan + Turkmen_Uzbekistan @ 2.885730

surbakhunWeesste
03-13-2015, 06:10 PM
my father's and mine

Bulut
03-13-2015, 06:38 PM
Population
Amerindian 0.16%
Ancestral_Altaic 2.48%
South_Central_Asian 14.29%
Arctic 1.38%
South_Indian 1.07%
Australoid 0.66%
Austronesian -
Caucasian 36.36%
Archaic_Human -
East_African 0.63%
East_Siberian 2.72%
European_Early_Farmers 11.53%
Khoisan -
Melano_Polynesian 0.26%
Archaic_African 1.06%
Near_East 8.98%
North_African 0.96%
Paleo_Siberian 1.89%
African_Pygmy 0.18%
South_East_Asian 1.90%
Subsaharian -
Tungus-Altaic 3.50%
European_Hunters_Gatherers 10.00%

Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Turk_Aydin @ 5.002108
2 Turk_Balikesir @ 6.713598
3 Turk_Istanbul @ 6.793200
4 Turk @ 7.329130
5 Turk_Adana @ 9.655854
6 Turk_Kayseri @ 9.984498
7 Azov_Greek @ 11.072625
8 Georgian_Jew @ 12.691505
9 Crimean_Tatar_Mountain @ 12.793129
10 Crimean_Tatar_Coast @ 13.972654
11 Cretan @ 14.108492
12 Azeri @ 14.575354
13 Kurd_North @ 14.647933
14 Cirkassian @ 14.852745
15 Uzbek_Tashkent @ 15.036784
16 Greek_Smyrna @ 15.177227
17 Stalskoe_Kumyk @ 15.479873
18 Syrian_Jew @ 15.759640
19 Kurd_South @ 16.151497
20 Assyrian_Iraqi @ 16.176567

Using 2 populations approximation:
1 50% Turk_Aydin +50% Turk_Aydin @ 5.002108


Using 3 populations approximation:
1 50% Greek +25% Turkmen_Uzbekistan +25% Uzbek_Tashkent @ 3.224803

Kaido
03-13-2015, 07:57 PM
Population
South_Central_Asian 33.02%
South_Indian 25.53%
Caucasian 20.65%
European_Hunters_Gatherers 7.84%
Ancestral_Altaic 5.20%
Tungus-Altaic 1.61%
Near_East 1.57%
Arctic 1.07%
Amerindian 1.02%
South_East_Asian 0.62%
Austronesian 0.50%
Khoisan 0.41%
Archaic_Human 0.31%
African_Pygmy 0.28%
East_Siberian 0.24%
Australoid 0.09%
Subsaharian 0.07%
East_African -
European_Early_Farmers -
Melano_Polynesian -
Archaic_African -
North_African -
Paleo_Siberian -

Rukha
03-14-2015, 12:50 AM
Kit M683104 (Afghan Hazara)

# Population Percent
1 Tungus-Altaic 20.25
2 South_Central_Asian 17.36
3 South_East_Asian 14.21
4 Caucasian 12.9
5 East_Siberian 7.25
6 European_Hunters_Gatherers 6.47
7 South_Indian 5.89
8 Ancestral_Altaic 4.74
9 Austronesian 3
10 European_Early_Farmers 2.46
11 Arctic 2.05
12 Near_East 1.2
13 Paleo_Siberian 0.96
14 Australoid 0.87
15 Amerindian 0.32
16 Archaic_African 0.06
17 Archaic_Human 0.02

Using 1 population approximation:
1 Uygur @ 4.466968
2 Hazara @ 6.193009
3 Uygur-Han @ 6.425982
4 Uzbek @ 8.865430
5 Tajik_Tajikistan @ 10.795231
6 Karakalpak @ 12.750441
7 Turkmen_Afghan @ 13.453513
8 Kazahs @ 13.814050
9 Turkmen_Uzbekistan @ 15.436490
10 Kazakh @ 19.125666
11 Tatar-Siberian @ 20.164406
12 Tajik_Afghan @ 20.461426
13 Kyrgyz_Bishkek @ 20.729958
14 Uzbek_Afghan @ 22.323896
15 Kyrgyz @ 23.908680
16 Spiti @ 24.495653
17 Crimean_Tatar_Step @ 26.697416
18 Nogai @ 27.814812
19 Bashkir @ 29.568695
20 Hakas @ 30.607704

Using 2 populations approximation:
1 50% Hazara +50% Uygur-Han @ 2.860135

Using 3 populations approximation:
1 50% Hazara +25% Uygur-Han +25% Uygur-Han @ 2.860135

Using 4 populations approximation:
1 Afghan_Pushtun + Mongola_China + Uygur + Uzbek @ 2.671150
2 Mongola_China + Pashtun_Afghani + Uygur-Han + Uzbek @ 2.837092
3 Hazara + Hazara + Uygur-Han + Uygur-Han @ 2.860135
4 Mongola_China + Pashtun_Afghani + Uygur + Uzbek @ 2.881691
5 Afghan_Pushtun + Buryat + Tajik_Yagnobi + Tu @ 2.895565
6 Mongola_China + Pashtun_Afghani + Turkmen_Afghan + Uygur-Han @ 2.914386
7 Kazahs + Mongola_China + Pashtun_Afghani + Turkmen_Uzbekistan @ 3.021547
8 Kazahs + Pashtun_Afghani + Turkmen_Uzbekistan + Xibo @ 3.028611
9 Mongol_Khalkha + Pashtun_Afghani + Uygur-Han + Uygur-Han @ 3.075534
10 Afghan_Pushtun + Mongola_China + Turkmen_Afghan + Uygur-Han @ 3.075535
11 Mongola_China + Tajik_Tajikistan + Uygur-Han + Uzbek_Afghan @ 3.089949
12 Mongola_China + Tajik_Afghan + Tajik_Tajikistan + Uygur-Han @ 3.102227

Kurd
03-14-2015, 12:53 AM
Another Iranian Baloch (Baloch3). He is more SE shifted than the other Iranian Baloch individuals that I posted. Noteworthy, is one of the highest SC Asian components from everyone posted so far (indicating proximity to the MDLP SC Asian reference), and Parsi again appearing as the closest surrogate.


http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20IRANIAN%20BALOCH_zpsfyugrjf7.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20IRANIAN%20BALOCH_zpsfyugrjf7.jpg.html )

Admix Results (sorted):
# Population Percent
1 South_Central_Asian 38.20
2 Caucasian 20.46
3 South_Indian 15.94
4 Near_East 10.46
5 European_Hunters_Gatherers 3.83
6 Ancestral_Altaic 3.02
7 North_African 2.38
8 Subsaharian 1.75
9 Archaic_African 1.04


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Parsi @ 10.319722
2 Afghan_Pushtun @ 10.637792
3 Pashtun_Afghani @ 11.074693
4 Pakistani_Pushtun @ 14.364246
5 Pathan @ 16.035828
6 Tajik_Pomiri_Ishkashim @ 16.466835
7 Uzbek_Afghan @ 16.479298
8 Iranian @ 18.467398
9 Tajik_Afghan @ 18.788801
10 Tajik_Pomiri_Shugnan @ 19.987513
11 Tajik_Pomiri_Rushan @ 20.065241
12 Burusho @ 20.384319
13 Makrani @ 20.827633
14 Jatt_Pahari @ 21.701597
15 Tajik_Yagnobi @ 22.345337
16 Punjabi_Gujjar @ 22.392242
17 Sindhi @ 22.729364
18 Balochi @ 23.581116
19 Kurd_South @ 24.605827
20 Jatt_Haryana @ 24.703547

Using 2 populations approximation:
1 50% Iranian +50% Sindhi @ 5.820503
Using 3 populations approximation:
1 50% Iranian +25% Makrani +25% Pakistani @ 4.414239

BalkanKiwi
03-14-2015, 01:04 AM
Population
Amerindian 0.33%
Ancestral_Altaic 4.46%
South_Central_Asian 5.65%
Australoid 0.64%
Caucasian 25.35%
European_Early_Farmers 24.52%
Melano_Polynesian 0.35%
North_African 2.15%
European_Hunters_Gatherers 36.55%

Kurd
03-14-2015, 03:32 PM
Based on some queries that I received, and I imagine there are other members interested, I will graph community member admixture overlap with Yamnaya. The MDLP k23 seems to be adequate for Yamnaya analysis since it is based on ~87,000 markers, and when I ran Yamnaya, it returned ~84,000. This will enable an apple to apples comparison (member to Yamnaya).

The Yamnaya reference/proxy will be created using the 2 Yamnaya profiles I have, I0443 and I0231. It would be helpful if someone could somehow get a gedmatch ID for Yamnaya I0429, since Davidsk has assigned it 0.999 Yamnaya related component in his K6 spreadsheet. I0429 would strengthen my Yamnaya proxy.

I looked into using Eurogenes calculators for this task, but the coverage for Yamnaya was low, since the Eurogenes calculators ( ANE k7 and K15) accuracy is based on 140,000+ markers. They are great calculators, but the accuracy would be limited for Yamnaya, since they returned about 50,000 to 60,000 SNPs. Thus it would be difficult to use them to do an apple to apples comparison.

Silesian
03-14-2015, 03:54 PM
Based on some queries that I received, ...............
Can you also add the H.G. R1a Karelian samples M652848

# Population Percent
1 European_Hunters_Gatherers 56.7
2 Ancestral_Altaic 25.7
3 Amerindian 9.97
4 South_Central_Asian 4.02
5 Arctic 3.05
6 African_Pygmy 0.22
7 Paleo_Siberian 0.19
8 Archaic_Human 0.08
9 East_Siberian 0.07

Kurd
03-14-2015, 03:58 PM
Can you also add the H.G. R1a Karelian samples M652848

# Population Percent
1 European_Hunters_Gatherers 56.7
2 Ancestral_Altaic 25.7
3 Amerindian 9.97
4 South_Central_Asian 4.02
5 Arctic 3.05
6 African_Pygmy 0.22
7 Paleo_Siberian 0.19
8 Archaic_Human 0.08
9 East_Siberian 0.07

Ok will do

Kurd
03-15-2015, 02:08 AM
Based on some queries that I received, and I imagine there are other members interested, I will graph community member admixture overlap with Yamnaya. The MDLP k23 seems to be adequate for Yamnaya analysis since it is based on ~87,000 markers, and when I ran Yamnaya, it returned ~84,000. This will enable an apple to apples comparison (member to Yamnaya).

The Yamnaya reference/proxy will be created using the 2 Yamnaya profiles I have, I0443 and I0231. It would be helpful if someone could somehow get a gedmatch ID for Yamnaya I0429, since Davidsk has assigned it 0.999 Yamnaya related component in his K6 spreadsheet. I0429 would strengthen my Yamnaya proxy.

I looked into using Eurogenes calculators for this task, but the coverage for Yamnaya was low, since the Eurogenes calculators ( ANE k7 and K15) accuracy is based on 140,000+ markers. They are great calculators, but the accuracy would be limited for Yamnaya, since they returned about 50,000 to 60,000 SNPs. Thus it would be difficult to use them to do an apple to apples comparison.

This is a friendly reminder that if anyone else wants to be included in this project they need to post their data under this thread by Sunday 3/15 0800 US Pacific Time, as I will not be able to do revisions or updates to add anyone posting after the deadline due to time constraints.

Also, for the data to be included it must be in the following format:

Population
Amerindian -
Ancestral_Altaic 3.61%
South_Central_Asian 7.61%
Arctic -
South_Indian -
Australoid 0.99%
Austronesian -
Caucasian 20.03%
Archaic_Human -
East_African -
East_Siberian -
European_Early_Farmers 27.81%
Khoisan -
Melano_Polynesian -
Archaic_African 0.93%
Near_East 1.16%
North_African 0.60%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 37.24%

Any different format, and I would have to manually enter, and it would literally takes me 10X as long to enter, and is subject to being rejected. Sorry, I just don't have that much time to spend on data entry.

EDIT: Those who have already posted, don't have to repost. I went ahead and manually entered your data. Any new posts will have to post in the format shown above.

BalkanKiwi
03-15-2015, 03:39 AM
For anyone who likes to look at Oracles I might as well include mine.

Using 1 population approximation:
1 German-Volga @ 3.277798
2 South_German @ 4.020086
3 Dutch @ 4.370359
4 Belgian @ 5.662082
5 Frisian @ 5.721272
6 Irish @ 6.471714
7 English @ 6.611763
8 Austrian @ 7.086512
9 North_German @ 7.705863
10 North_European @ 7.717642
11 Dane @ 8.460835
12 English_Kent_GBR @ 8.958603
13 Norwegian_East @ 9.030949
14 German_East @ 9.284139
15 CEU @ 9.539107
16 Serb_Serbia @ 9.638930
17 English_Cornwall_GBR @ 9.687665
18 Swede @ 10.019649
19 British @ 10.212126
20 Welsh @ 10.303917

Using 2 populations approximation:
1 50% Scottish_Argyll_Bute_GBR +50% Serb_Serbia @ 2.370529


Using 3 populations approximation:
1 50% Dutch +25% Frisian +25% Montenegrian @ 2.053689


Using 4 populations approximation:
1 Bulgarian + Dutch + Dutch + Frisian @ 1.824983
2 Dutch + German-Volga + Irish + Serb_Serbia @ 1.826487
3 Dutch + Frisian + German-Volga + Serb_Serbia @ 1.883198
4 Bulgarian + Dutch + Dutch + North_European @ 1.926490
5 Belgian + Bulgarian + Dutch + Dutch @ 1.950422
6 Bulgarian + Dutch + Dutch + Irish @ 1.953127
7 Bulgarian + Dane + Frisian + Irish @ 1.977519
8 Bulgarian + Dutch + Dutch + English @ 1.981946
9 Bulgarian + Dutch + Frisian + Norwegian_East @ 1.995352
10 Dutch + English + German-Volga + Serb_Serbia @ 1.999415
11 Bulgarian + Dane + Frisian + Frisian @ 2.000729
12 Dutch + German-Volga + North_European + Serb_Serbia @ 2.028015
13 Dutch + Dutch + Frisian + Montenegrian @ 2.053689
14 Dutch + Dutch + Irish + Montenegrian @ 2.062840
15 Bulgarian + Dutch + Dutch + English_Kent_GBR @ 2.080451
16 Bulgarian + Dutch + Irish + Norwegian_East @ 2.088581
17 Bosnian + Frisian + German-Volga + Irish @ 2.093210
18 Bulgarian + Dutch + Frisian + Frisian @ 2.097251
19 Bulgarian + Frisian + German-Volga + Norwegian_East @ 2.100228
20 Bosnian + German-Volga + Irish + Irish @ 2.102215

Táltos
03-15-2015, 03:58 AM
Thanks Kurd I was just going to repost mine. Sorry about how I had my format.

CelticGerman
03-15-2015, 06:37 AM
Here is mine:

Population
Amerindian -
Ancestral_Altaic 4.84%
South_Central_Asian 5.18%
Arctic -
South_Indian -
Australoid 0.10%
Austronesian -
Caucasian 24.23%
Archaic_Human -
East_African -
East_Siberian -
European_Early_Farmers 25.57%
Khoisan -
Melano_Polynesian 0.31%
Archaic_African -
Near_East -
North_African -
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian 0.19%
Tungus-Altaic -
European_Hunters_Gatherers 39.58%

Varun R
03-15-2015, 07:33 AM
Population
Amerindian -
Ancestral_Altaic 3.74%
South_Central_Asian 32.09%
Arctic 0.32%
South_Indian 45.85%
Australoid 0.39%
Austronesian -
Caucasian 7.11%
Archaic_Human -
East_African -
East_Siberian 0.95%
European_Early_Farmers 0.72%
Khoisan 0.53%
Melano_Polynesian -
Archaic_African 0.53%
Near_East -
North_African -
Paleo_Siberian -
African_Pygmy -
South_East_Asian 2.10%
Subsaharian 0.14%
Tungus-Altaic 0.51%
European_Hunters_Gatherers 4.94%

Least-squares method.

Using 1 population approximation:
1 GujaratiB_GIH @ 3.770789
2 Hindi @ 4.103536
3 Brahmin_Tamil @ 4.256459
4 Vaish @ 4.359940
5 Kshatriya @ 4.662208
6 Tiwari @ 5.035013
7 Marathi @ 5.111001
8 Brahmins_UP @ 6.443657
9 TN_Brahmin @ 6.613381
10 Meghawal @ 8.329546
11 GujaratiC_GIH @ 8.971700
12 Jatt_Muslim @ 9.015787
13 Pakistani @ 10.230445
14 GujaratiA_GIH @ 11.045025
15 Punjabi_Lahore_PJL @ 11.166000
16 Telugu_Kannada @ 11.270144
17 Bengali @ 11.520662
18 Dharkar @ 13.454458
19 Tamil_Singapore @ 13.763741
20 Srivastava @ 13.883783

Using 2 populations approximation:
1 50% Brahmin_Tamil +50% Tiwari @ 2.450356

*This is the most accurate 2 population approximation out of all calculators I've used.

Shuffle
03-15-2015, 04:03 PM
# Population Percent
1 European_Hunters_Gatherers 33.67
2 European_Early_Farmers 28.26
3 Caucasian 24.72
4 South_Central_Asian 4.63
5 Ancestral_Altaic 2.78
6 North_African 2.71


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 German-Volga @ 4.241796
2 South_German @ 5.144444
3 Belgian @ 5.466467
4 English @ 6.196717
5 Frisian @ 6.208614
6 Irish @ 6.481298
7 English_Kent_GBR @ 6.991228
8 North_European @ 7.444989
9 English_Cornwall_GBR @ 7.665088
10 Dutch @ 7.768205
11 Welsh @ 8.183524
12 CEU @ 8.303096
13 British @ 8.901015
14 Italian_North @ 9.967720
15 French @ 10.299084
16 Scottish_Argyll_Bute_GBR @ 10.544416
17 Austrian @ 10.673048
18 North_German @ 11.093665
19 Orcadian @ 11.202626
20 Norwegian_East @ 11.217471

This is mine, if something else is needed, tell me. ;-)

fil
03-15-2015, 04:42 PM
Here is a friend of mine. He is the same ethnicity as me.

Population
Amerindian -
Ancestral_Altaic 1.38%
South_Central_Asian 32.74%
Arctic -
South_Indian 46.97%
Australoid 0.08%
Austronesian 1.01%
Caucasian 7.08%
Archaic_Human 0.61%
East_African -
East_Siberian 1.09%
European_Early_Farmers -
Khoisan -
Melano_Polynesian 0.67%
Archaic_African -
Near_East 2.14%
North_African 2.29%
Paleo_Siberian 1.92%
African_Pygmy -
South_East_Asian 1.50%
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 0.53%

Kurd
03-15-2015, 05:01 PM
This project will result in multiple charts, graphs, and PCAs being uploaded over the next couple of weeks. I recommend that you regularly check for updates.

The first 2 graphs I am posting show the community’s admixture overlap with Yamnaya and Karelian. I feel good about using the MDLP k23b calculator on the ancient genomes, because the calculator had over 95% marker overlap, and thus the results can be relied upon in comparisons with members’ admixture.

The results are consistent with Davidski’s findings at http://eurogenes.blogspot.com/2015/03/yamnaya-related-ancestry-proportions-in.html. Although I believe that the MDLP k23 is much better suited for ancient admixture, because the Eurogenes calculators I checked, ANE k7 and K15, had an overlap of only about 35% with the ancient genomes.

Similar to my findings, his findings also showed NE Europeans as having very high Yamnaya related ancestry. This is reflected in my graph, showing Salkin, Silesian, Celticgerman in the lead.

With Asians, Davidski showed that Tajiks, Kalash, and Pashtuns had the highest levels. My calculations were in agreement, in that Rukha (50%Tajik/50% Pashtun) had the highest score, followed by other Pashtun/Pathans, Kandhari Durrani, AdamYZ, Sein…, followed by Iranians and Kurds, led by DMXX.
Of significance, the analysis verifies that a major event occurred to change the genetic landscape of Europe between the early and late Neolithic period. In Hungary, BR2’s Yamnaya levels are more than double NE1’s.

I added Karelian HG , because it adds another dimension to the analysis. Significantly, at around 7ky LBK Stuttgart’s admixture similarity to Karelian HG is at 0%. NE1’s similarity at 7ky is only about 5%. Yet, at 3.2ky, BR2’s similarity to Karelian HG is almost 40%, and modern Europeans & Mediterraneans (community members) are at 30 to 55%. Again, Salkin, Silesian, and Celticgerman are in the lead here.

Another observation is that in Europe, Karelian HG levels shot up more than Yamnaya levels from early to late Neolithic. This points to events taking place that brought significant levels of Karelian to Europe.

http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20OVERLAP%20KARELIAN_zpsi4t7hivh.jpg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20OVERLAP%20KARELIAN_zpsi4t7hivh.jpg.htm l)

http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20OVERLAP%20YAMNAYA_zpsqpafyukt.jpg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20OVERLAP%20YAMNAYA_zpsqpafyukt.jpg.html )

vettor
03-15-2015, 06:18 PM
This project will result in multiple charts, graphs, and PCAs being uploaded over the next couple of weeks. I recommend that you regularly check for updates.

The first 2 graphs I am posting show the community’s admixture overlap with Yamnaya and Karelian. I feel good about using the MDLP k23b calculator on the ancient genomes, because the calculator had over 95% marker overlap, and thus the results can be relied upon in comparisons with members’ admixture.

The results are consistent with Davidski’s findings at http://eurogenes.blogspot.com/2015/03/yamnaya-related-ancestry-proportions-in.html. Although I believe that the MDLP k23 is much better suited for ancient admixture, because the Eurogenes calculators I checked, ANE k7 and K15, had an overlap of only about 35% with the ancient genomes.

Similar to my findings, his findings also showed NE Europeans as having very high Yamnaya related ancestry. This is reflected in my graph, showing Salkin, Silesian, Celticgerman in the lead.

With Asians, Davidski showed that Tajiks, Kalash, and Pashtuns had the highest levels. My calculations were in agreement, in that Rukha (50%Tajik/50% Pashtun) had the highest score, followed by other Pashtun/Pathans, Kandhari Durrani, AdamYZ, Sein…, followed by Iranians and Kurds, led by DMXX.
Of significance, the analysis verifies that a major event occurred to change the genetic landscape of Europe between the early and late Neolithic period. In Hungary, BR2’s Yamnaya levels are more than double NE1’s.

I added Karelian HG , because it adds another dimension to the analysis. Significantly, at around 7ky LBK Stuttgart’s admixture similarity to Karelian HG is at 0%. NE1’s similarity at 7ky is only about 5%. Yet, at 3.2ky, BR2’s similarity to Karelian HG is almost 40%, and modern Europeans & Mediterraneans (community members) are at 30 to 55%. Again, Salkin, Silesian, and Celticgerman are in the lead here.

Another observation is that in Europe, Karelian HG levels shot up more than Yamnaya levels from early to late Neolithic. This points to events taking place that brought significant levels of Karelian to Europe.

http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20OVERLAP%20KARELIAN_zpsi4t7hivh.jpg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20OVERLAP%20KARELIAN_zpsi4t7hivh.jpg.htm l)

http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20OVERLAP%20YAMNAYA_zpsqpafyukt.jpg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20OVERLAP%20YAMNAYA_zpsqpafyukt.jpg.html )

thank for info

well, well, in your second chart, Agamemnon, Sapporo and myself are the "bastards" in this mix............neither here nor there ...........maybe what the ancient anatolians where before being turkenized!

Kurd
03-15-2015, 06:32 PM
thank for info

well, well, in your second chart, Agamemnon, Sapporo and myself are the "bastards" in this mix............neither here nor there ...........maybe what the ancient anatolians where before being turkenized!

Im not sure what you meant by that. You are aware that Sapporo has Indian ancestry. I made a mistake in his coloring

vettor
03-15-2015, 06:35 PM
Im not sure what you meant by that. You are aware that Sapporo has Indian ancestry. I made a mistake in his coloring

there is no intent in my comment.............only that these 3 europeans stand out form the other europeans.

I enjoy your good work :)

rms2
03-15-2015, 06:48 PM
I guess I am too late to this thread to really get in on it, but here are my MDLP K23b results anyway.

4051

rms2
03-15-2015, 06:55 PM
I guess I am too late to this thread to really get in on it, but here are my MDLP K23b results anyway.

4051

I guess it's impossible to copy and paste the stats the way I presented them (as a picture), so here they are copied and pasted:

Population
Amerindian 1.31%
Ancestral_Altaic 3.60%
South_Central_Asian 5.50%
Arctic -
South_Indian -
Australoid -
Austronesian -
Caucasian 20.98%
Archaic_Human -
East_African -
East_Siberian -
European_Early_Farmers 28.22%
Khoisan -
Melano_Polynesian 0.14%
Archaic_African -
Near_East 0.44%
North_African -
Paleo_Siberian -
African_Pygmy -
South_East_Asian 0.62%
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 39.15%

seferhabahir
03-15-2015, 07:12 PM
I'm also late to the party.

Here they are anyway in case you decide to do another iteration on the charts.

Population
Amerindian -
Ancestral_Altaic 0.84%
South_Central_Asian 6.97%
Arctic -
South_Indian -
Australoid 0.37%
Austronesian 0.63%
Caucasian 36.32%
Archaic_Human -
East_African 1.67%
East_Siberian 1.52%
European_Early_Farmers 19.13%
Khoisan -
Melano_Polynesian 0.16%
Archaic_African -
Near_East 13.35%
North_African 5.44%
Paleo_Siberian 0.91%
African_Pygmy -
South_East_Asian 0.47%
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 12.20%

Kurd
03-15-2015, 07:49 PM
This project will result in multiple charts, graphs, and PCAs being uploaded over the next couple of weeks. I recommend that you regularly check for updates.

The first 2 graphs I am posting show the community’s admixture overlap with Yamnaya and Karelian. I feel good about using the MDLP k23b calculator on the ancient genomes, because the calculator had over 95% marker overlap, and thus the results can be relied upon in comparisons with members’ admixture.

The results are consistent with Davidski’s findings at http://eurogenes.blogspot.com/2015/03/yamnaya-related-ancestry-proportions-in.html. Although I believe that the MDLP k23 is much better suited for ancient admixture, because the Eurogenes calculators I checked, ANE k7 and K15, had an overlap of only about 35% with the ancient genomes.

Similar to my findings, his findings also showed NE Europeans as having very high Yamnaya related ancestry. This is reflected in my graph, showing Salkin, Silesian, Celticgerman in the lead.

With Asians, Davidski showed that Tajiks, Kalash, and Pashtuns had the highest levels. My calculations were in agreement, in that Rukha (50%Tajik/50% Pashtun) had the highest score, followed by other Pashtun/Pathans, Kandhari Durrani, AdamYZ, Sein…, followed by Iranians and Kurds, led by DMXX.
Of significance, the analysis verifies that a major event occurred to change the genetic landscape of Europe between the early and late Neolithic period. In Hungary, BR2’s Yamnaya levels are more than double NE1’s.

I added Karelian HG , because it adds another dimension to the analysis. Significantly, at around 7ky LBK Stuttgart’s admixture similarity to Karelian HG is at 0%. NE1’s similarity at 7ky is only about 5%. Yet, at 3.2ky, BR2’s similarity to Karelian HG is almost 40%, and modern Europeans & Mediterraneans (community members) are at 30 to 55%. Again, Salkin, Silesian, and Celticgerman are in the lead here.

Another observation is that in Europe, Karelian HG levels shot up more than Yamnaya levels from early to late Neolithic. This points to events taking place that brought significant levels of Karelian to Europe.

http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20OVERLAP%20KARELIAN_zpsi4t7hivh.jpg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20OVERLAP%20KARELIAN_zpsi4t7hivh.jpg.htm l)

http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20OVERLAP%20YAMNAYA_zpsqpafyukt.jpg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20OVERLAP%20YAMNAYA_zpsqpafyukt.jpg.html )

The following graph shows the community's admixture overlap with 45ky Ust-Ishim from Siberia. South Asians score the highest, with Dr McNinja at around 50%. Iranians/Kurds are next, with my score at 40%.



http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20OVERLAP%20UST%20ISHIM_zpsl9xnw7mx.jpg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20OVERLAP%20UST%20ISHIM_zpsl9xnw7mx.jpg. html)

Humanist
03-15-2015, 07:59 PM
If you wish to include them at any point, here are my values, in the requested format*:

Population
Amerindian 0.64%
Ancestral_Altaic 1.39%
South_Central_Asian 19.68%
Arctic -
South_Indian 1.64%
Australoid -
Austronesian -
Caucasian 42.58%
Archaic_Human 0.23%
East_African -
East_Siberian -
European_Early_Farmers 8.49%
Khoisan -
Melano_Polynesian 0.64%
Archaic_African -
Near_East 18.91%
North_African 4.55%
Paleo_Siberian 0.55%
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic 0.41%
European_Hunters_Gatherers 0.29%

* I previously posted my values (http://www.anthrogenica.com/showthread.php?4009-MDLP-K23b-Project-Part-2-of-2-Community-PCAs-graphs-amp-charts&p=73991&viewfull=1#post73991), but they were in the incorrect format.

Kurd
03-15-2015, 08:18 PM
If you wish to include them at any point, here are my values, in the requested format*:

Population
Amerindian 0.64%
Ancestral_Altaic 1.39%
South_Central_Asian 19.68%
Arctic -
South_Indian 1.64%
Australoid -
Austronesian -
Caucasian 42.58%
Archaic_Human 0.23%
East_African -
East_Siberian -
European_Early_Farmers 8.49%
Khoisan -
Melano_Polynesian 0.64%
Archaic_African -
Near_East 18.91%
North_African 4.55%
Paleo_Siberian 0.55%
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic 0.41%
European_Hunters_Gatherers 0.29%

* I previously posted my values (http://www.anthrogenica.com/showthread.php?4009-MDLP-K23b-Project-Part-2-of-2-Community-PCAs-graphs-amp-charts&p=73991&viewfull=1#post73991), but they were in the incorrect format.

I am not sure how you fell through the cracks. Since it was my mistake, and you had posted before the deadline, I revised my last graph to include your data.

The following graph shows the community's admixture overlap with 45ky Ust-Ishim from Siberia. South Asians score the highest, with Dr McNinja at around 50%. Iranians/Kurds are next, with my score at 40%.



http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20OVERLAP%20UST%20ISHIM_zpsgu54d5fe.jpg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20OVERLAP%20UST%20ISHIM_zpsgu54d5fe.jpg. html)

Calamus
03-15-2015, 08:21 PM
My results

Population
Amerindian -
Ancestral_Altaic 5.52%
South_Central_Asian 4.14%
Arctic 1.10%
South_Indian -
Australoid 0.82%
Austronesian -
Caucasian 22.58%
Archaic_Human 0.05%
East_African -
East_Siberian -
European_Early_Farmers 23.16%
Khoisan 0.38%
Melano_Polynesian 0.07%
Archaic_African 0.05%
Near_East 0.24%
North_African -
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 41.89%

Silesian
03-15-2015, 09:46 PM
As per request, proper format
Poland-My results

Population
Amerindian -
Ancestral_Altaic 2.91%
South_Central_Asian 3.87%
Arctic -
South_Indian 0.66%
Australoid 0.15%
Austronesian 0.43%
Caucasian 27.23%
Archaic_Human -
East_African -
East_Siberian 0.64%
European_Early_Farmers 13.91%
Khoisan -
Melano_Polynesian 0.33%
Archaic_African -
Near_East 0.50%
North_African -
Paleo_Siberian 0.74%
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 48.61%

Russian1 sample

Population
Amerindian -
Ancestral_Altaic 2.66%
South_Central_Asian 3.62%
Arctic 0.47%
South_Indian 0.08%
Australoid -
Austronesian 1.11%
Caucasian 29.41%
Archaic_Human 0.13%
East_African -
East_Siberian 0.34%
European_Early_Farmers 11.39%
Khoisan -
Melano_Polynesian 0.24%
Archaic_African -
Near_East -
North_African -
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic 1.20%
European_Hunters_Gatherers 49.37%




Russian2 sample

Population
Amerindian 0.66%
Ancestral_Altaic 6.57%
South_Central_Asian 1.46%
Arctic 1.66%
South_Indian -
Australoid -
Austronesian 1.48%
Caucasian 17.17%
Archaic_Human 0.12%
East_African -
East_Siberian 3.69%
European_Early_Farmers 13.29%
Khoisan 0.41%
Melano_Polynesian 1.09%
Archaic_African -
Near_East -
North_African 0.39%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 52.02%

Baltic1 sample
Population
Amerindian 0.46%
Ancestral_Altaic 1.99%
South_Central_Asian 0.29%
Arctic 0.18%
South_Indian 0.95%
Australoid 0.26%
Austronesian 0.54%
Caucasian 29.36%
Archaic_Human -
East_African -
East_Siberian 0.58%
European_Early_Farmers 9.02%
Khoisan -
Melano_Polynesian -
Archaic_African 0.12%
Near_East 1.05%
North_African -
Paleo_Siberian -
African_Pygmy -
South_East_Asian 0.27%
Subsaharian 0.26%
Tungus-Altaic 0.43%
European_Hunters_Gatherers 54.24%

Dr_McNinja
03-15-2015, 10:16 PM
I am not sure how you fell through the cracks. Since it was my mistake, and you had posted before the deadline, I revised my last graph to include your data.

The following graph shows the community's admixture overlap with 45ky Ust-Ishim from Siberia. South Asians score the highest, with Dr McNinja at around 50%. Iranians/Kurds are next, with my score at 40%.



http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20OVERLAP%20UST%20ISHIM_zpsgu54d5fe.jpg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20OVERLAP%20UST%20ISHIM_zpsgu54d5fe.jpg. html)
That's pretty interesting.

I wonder if the noise South Asians get in some calculators (like Oceanian/East Eurasian in K8) isn't just derived from this really old population (or something similar). It gets high near the Himalayas too (HRP0402 doesn't have high east eurasian like me or parasar (bihar), but is a "west pahari" origin from northwest punjab near the south end of Pakistan's azad kashmir).

Sein also had some of this type of noise.

Kurd
03-16-2015, 12:28 AM
That's pretty interesting.

I wonder if the noise South Asians get in some calculators (like Oceanian/East Eurasian in K8) isn't just derived from this really old population (or something similar). It gets high near the Himalayas too (HRP0402 doesn't have high east eurasian like me or parasar (bihar), but is a "west pahari" origin from northwest punjab near the south end of Pakistan's azad kashmir).

Sein also had some of this type of noise.

I was pleasantly surprised that we had this much overlap with the oldest sequenced human genome (~45ky). I also found it interesting that I an trace back to his lineages both paternally and maternally, LT and R respectively.

With regards to the noise, it is possible, because Ust is not that long, relatively speaking, after out-of-Africa, right around or before the East Eurasian West Eurasian split. Maybe Parasar can also comment.

Táltos
03-16-2015, 03:25 AM
That's pretty interesting.

I wonder if the noise South Asians get in some calculators (like Oceanian/East Eurasian in K8) isn't just derived from this really old population (or something similar). It gets high near the Himalayas too (HRP0402 doesn't have high east eurasian like me or parasar (bihar), but is a "west pahari" origin from northwest punjab near the south end of Pakistan's azad kashmir).

Sein also had some of this type of noise.

Out of the Europeans that participated, it looks like I have the most overlap with Ust-Ishim. I scored Melano_Polynesian 1.00 in the K23b calculator. Wonder if that's why?

Thank you Kurd for your project. This is really cool!

Kurd
03-16-2015, 01:08 PM
The following graph shows the community's admixture overlap with 45ky Ust-Ishim from Siberia. South Asians score the highest, with Dr McNinja at around 50%. Iranians/Kurds are next, with my score at 40%.



http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20OVERLAP%20UST%20ISHIM_zpsl9xnw7mx.jpg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20OVERLAP%20UST%20ISHIM_zpsl9xnw7mx.jpg. html)

Ust-Ishim is more similar to East Asians than to Europeans, since his genome shares more derived alleles with modern East Asians.

MDLP k23's closest proxies for Ust-Ishim are:

1 Ayta_AE @ 24.783888
2 Nepalese @ 27.270012
3 Kensiu @ 29.925926
4 Tamil_Singapore @ 29.993927
5 Cochin_Jew @ 30.449663
6 Onge @ 31.218340
7 Pahari @ 31.865376
8 Mumbai_Jew @ 33.207047
9 Dhaka_mixed_popul @ 33.615238
10 Hindi @ 34.281342
11 Uygur-Han @ 34.523560
12 Jatt_Haryana @ 34.819286
13 Tiwari @ 35.455795
14 Pakistani_Pushtun @ 35.725113
15 Marathi @ 35.728867
16 Jatt_Muslim @ 35.853012
17 Mamanawa @ 36.317513
18 Vaish @ 36.851608
19 Brahmins_UP @ 36.884640
20 Spiti @ 37.032146

When his genome was compared to 8ky La Brana and 24ky MA-1, it appeared that he shared similar amounts of derived alleles with them, as he did with modern East Asians.

The following quote is taken from Dienekes's blog at http://dienekes.blogspot.com/2014/10/high-coverage-genome-from-45000-year.html:

"This suggests that the population to which the Ust’-Ishim individual belonged diverged from the ancestors of present-day West Eurasian and East Eurasian populations before—or simultaneously with—their divergence from each other. The finding that the Ust’-Ishim individual is equally closely related to present-day Asians and to 8,000- to 24,000-year-old individuals from western Eurasia, but not to present-day Europeans, is compatible with the hypothesis that present-day Europeans derive some of their ancestry from a population that did not participate in the initial dispersals of modern humans into Europe and Asia"

Anabasis
03-16-2015, 02:14 PM
I know i am late but. My results are here.

Amerindian -
Ancestral_Altaic -
South_Central_Asian 17.90%
Arctic -
South_Indian 0.71%
Australoid -
Austronesian -
Caucasian 55.82%
Archaic_Human 0.23%
East_African -
East_Siberian -
European_Early_Farmers 7.44%
Khoisan -
Melano_Polynesian 1.06%
Archaic_African -
Near_East 12.12%
North_African 3.61%
Paleo_Siberian -
African_Pygmy 0.16%
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 0.95%

seferhabahir
03-16-2015, 04:58 PM
Out of the Europeans that participated, it looks like I have the most overlap with Ust-Ishim. I scored Melano_Polynesian 1.00 in the K23b calculator. Wonder if that's why?

I'm not on the chart, but wonder where I would show up if I was, presuming I would be categorized as European, probably on the right in the middle of the Armenians and Europeans for Ust-Ishim, but I would be most interested in the Yamnaya and Karelian placement. My top four scores have obvious affinity to Mediterranean. I hope Kurd can find time later in the month to rerun the plots with the newest submissions.

1 Ashkenazi_Jew @ 5.489746
2 Cretan @ 5.749528
3 Sicilian_West @ 6.405828
4 Sicilian_East @ 6.957144

Táltos
03-16-2015, 07:08 PM
I'm not on the chart, but wonder where I would show up if I was, presuming I would be categorized as European, probably on the right in the middle of the Armenians and Europeans for Ust-Ishim, but I would be most interested in the Yamnaya and Karelian placement. My top four scores have obvious affinity to Mediterranean. I hope Kurd can find time later in the month to rerun the plots with the newest submissions.

1 Ashkenazi_Jew @ 5.489746
2 Cretan @ 5.749528
3 Sicilian_West @ 6.405828
4 Sicilian_East @ 6.957144

Yes I hope he can update for the new submissions too. For me I'm the 8th on both charts for overlapping with Yamnaya and Karelian. Here's my top four single population sharing scores from MDLP K23b if that helps at all.

1 Austrian @ 4.205538
2 Hungarian_Budapest @ 4.917678
3 Hungarian @ 5.610407
4 Serb_Serbia @ 5.870380

Agamemnon
03-16-2015, 07:34 PM
I'm not on the chart, but wonder where I would show up if I was, presuming I would be categorized as European, probably on the right in the middle of the Armenians and Europeans for Ust-Ishim, but I would be most interested in the Yamnaya and Karelian placement. My top four scores have obvious affinity to Mediterranean. I hope Kurd can find time later in the month to rerun the plots with the newest submissions.

1 Ashkenazi_Jew @ 5.489746
2 Cretan @ 5.749528
3 Sicilian_West @ 6.405828
4 Sicilian_East @ 6.957144

Particularly interesting because I also get 50% Cretan as my best fit with MDLP k23b, it seems that Cretans are more similar to AJs than Sicilians or South Italians.

everest59
03-16-2015, 10:57 PM
I know that this is after the deadline, but kit M821342 could be interesting. This is a zombie of Harappa's "Baloch" component that I uploaded to Gedmatch.

1 South_Central_Asian 75.16
2 Caucasian 8.60
3 Ancestral_Altaic 7.61
4 South_Indian 6.58
5 Near_East 1.32

parasar
03-17-2015, 12:46 AM
...

MDLP k23's closest proxies for Ust-Ishim are:

1 Ayta_AE @ 24.783888
...

Ust-Ishim (~45000ybp) is close to the negrito types (unadmixed with Denisovan) and also close to the NO node and M526 overall.
If notice in Fig S3 here - http://genome.cshlp.org/content/suppl/2015/02/18/gr.186684.114.DC1/Supplemental_Figures.pdf - we have under K# (~50000ybp) the following: Agta O, Agta P, and Aeta MS.

seferhabahir
03-17-2015, 01:14 AM
Particularly interesting because I also get 50% Cretan as my best fit with MDLP k23b, it seems that Cretans are more similar to AJs than Sicilians or South Italians.

I think this depends on who's autosomal DNA got tested. We discussed this a little bit in the K23b results thread where John Doe (also AJ) posted his K23b ethnicity results as closer to Sicilian than to Cretan. Looks like there are some different reference populations for AJ as well in the Oracle calculator. I know that my autosomal DNA is typical for AJ, but the Cretan scoring higher than Sicilian is still interesting.

John Doe using 1 population approximation:
1 Ashkenazi_Jew @ 2.929181
2 Sicilian_East @ 3.546232
3 Sicilian_West @ 3.845939
4 Sicilian_Siracusa @ 3.939226
5 Ashkenazi @ 4.350750
6 Sicilian_Trapani @ 5.250281
7 Sicilian_Agrigento @ 5.379285
8 Romanian_Jew @ 5.465404
9 Maltese @ 6.141469
10 Cretan @ 6.226387

It may depend on how many Mizrachi remnants there are in the DNA. I know I have a bit which might move me somewhat eastward, hence closer to Crete than Sicily as compare to John Doe above. Nat Geo, with their lack of any AJ reference populations, gave me Sardinia and Iran, which was somewhat meaningless in terms of pinpointing something specific. Nat Geo does not consider AJ to have an identifiable geographic location.

Agamemnon
03-17-2015, 02:33 AM
I think this depends on who's autosomal DNA got tested. We discussed this a little bit in the K23b results thread where John Doe (also AJ) posted his K23b ethnicity results as closer to Sicilian than to Cretan. Looks like there are some different reference populations for AJ as well in the Oracle calculator. I know that my autosomal DNA is typical for AJ, but the Cretan scoring higher than Sicilian is still interesting.

John Doe using 1 population approximation:
1 Ashkenazi_Jew @ 2.929181
2 Sicilian_East @ 3.546232
3 Sicilian_West @ 3.845939
4 Sicilian_Siracusa @ 3.939226
5 Ashkenazi @ 4.350750
6 Sicilian_Trapani @ 5.250281
7 Sicilian_Agrigento @ 5.379285
8 Romanian_Jew @ 5.465404
9 Maltese @ 6.141469
10 Cretan @ 6.226387

It may depend on how many Mizrachi remnants there are in the DNA. I know I have a bit which might move me somewhat eastward, hence closer to Crete than Sicily as compare to John Doe above. Nat Geo, with their lack of any AJ reference populations, gave me Sardinia and Iran, which was somewhat meaningless in terms of pinpointing something specific. Nat Geo does not consider AJ to have an identifiable geographic location.

Possible, it would make sense for me as well since I suspect Mizrahi ancestry in my paternal grandmother's family (they were Crimean Jews), and this would help explain all the Uzbek Jewish matches (as well as the Syrian & Iranian Jewish ones) in my RF.

Kurd
03-17-2015, 02:34 AM
I know that this is after the deadline, but kit M821342 could be interesting. This is a zombie of Harappa's "Baloch" component that I uploaded to Gedmatch.

1 South_Central_Asian 75.16
2 Caucasian 8.60
3 Ancestral_Altaic 7.61
4 South_Indian 6.58
5 Near_East 1.32

SC Asian seems high for the zombie to be based on actual references. For meaningful results, it should be based on actual references. Do you know what steps were taken to create it? Feel free to PM me if you like.

Kurd
03-17-2015, 02:39 AM
Has anyone seen an FST table for MDLP k23? It would be helpful for me to use as a reference when I create the components for the PCA.

Dr_McNinja
03-17-2015, 03:05 AM
SC Asian seems high for the zombie to be based on actual references. For meaningful results, it should be based on actual references. Do you know what steps were taken to create it? Feel free to PM me if you like.

It's just a cluster/component created in an unsupervised run of Admixture. It's the Baloch/Brahui/Makrani cluster where they hit 50-60% of the component at max.

Kurd
03-17-2015, 01:05 PM
It's just a cluster/component created in an unsupervised run of Admixture. It's the Baloch/Brahui/Makrani cluster where they hit 50-60% of the component at max.

This is the MDLP oracle for the "zombie":

Admix Results (sorted):


# Population Percent
1 South_Central_Asian 75.16
2 Caucasian 8.60
3 Ancestral_Altaic 7.61
4 South_Indian 6.58
5 Near_East 1.32


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Kalash @ 13.634830
2 Brahui @ 17.620451
3 Balochi @ 21.447117
4 Makrani @ 24.130192
5 Sindhi @ 39.364365
6 Pashtun_Afghani @ 40.144093
7 Pathan @ 40.551128
8 Burusho @ 41.807949
9 Punjabi_Gujjar @ 42.194130
10 Tajik_Pomiri_Ishkashim @ 42.330921
11 Afghan_Pushtun @ 43.496132
12 Tajik_Pomiri_Shugnan @ 44.822605
13 Tajik_Pomiri_Rushan @ 45.796021
14 Uzbek_Afghan @ 47.342617
15 Jatt_Pahari @ 47.564907
16 Tajik_Yagnobi @ 48.251320
17 GujaratiA_GIH @ 48.911804
18 Pakistani @ 48.950233
19 Pakistani_Pushtun @ 49.574814
20 Parsi @ 49.724422

Using 2 populations approximation:
1 50% Kalash +50% Kalash @ 13.634830


Using 3 populations approximation:
1 50% Kalash +25% Kalash +25% Kalash @ 13.634830

I think it will be more informative to compare to a Kalash reference, comprised of a mean of actual Kalash individuals, which will make a better proxy for Kalash. Alternatively, actual ancient S or SC DNA would be ideal.

On a different note, do you have any data for Parsis. I am interested in doing a comparison with it, since it shows up as one of the best proxies for the Iranian Baloch individuals, myself, and Jesus, and it would be nice to have them included on the PCA along with Kalash.

Raskolnikov
03-18-2015, 04:00 PM
A bit late but I just saw this now, here's my results.

Population
Amerindian 0.24%
Ancestral_Altaic 6.97%
South_Central_Asian 34.77%
Arctic 1.16%
South_Indian 31.09%
Australoid -
Austronesian 0.68%
Caucasian 14.14%
Archaic_Human -
East_African -
East_Siberian 0.52%
European_Early_Farmers 1.72%
Khoisan -
Melano_Polynesian -
Archaic_African 0.38%
Near_East 2.24%
North_African 0.12%
Paleo_Siberian -
African_Pygmy -
South_East_Asian 2.23%
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 3.75%

I'm Pakistani Punjabi.

Kurd
03-21-2015, 10:49 AM
First, a little on Principal Component Analysis (PCA), also known as multi-variate analysis for those not familiar with it. Untill now, I have used 2D charts to plot members' admixture variations. 2D charts work well as long as the number of components is low. Since I have been using low k calculators so far, this has worked well, because I have been able to capture a high percentage of the members' admixture with a couple of charts, with each chart capturing 2 components.

With higher k calculators, I need to be able to reduce multiple dimensions (dimensionality of multiple components) down to a few dimensions or components. PCA does this by reducing the dimensionality of of multiple variables down to a few principal components. The PCA algorithm minimizes the loss of information/variability when the dimensionality of the calculator is reduced.

With the MDLP k23, I have been able to capture about 62% of the member's genetic variation with 2 prinicipal components. With the 1st 4 principal components, I am at around 90%. I may post a bi-plot showing PC3 and PC4 later. The software has weighted the components as follows:

http://i1377.photobucket.com/albums/ah42/dil88888/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg (http://s1377.photobucket.com/user/dil88888/media/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg.html)

For legibility reasons, I have de-activated the Eigen vector layer from the PCA bi-plot. For those interested, the following are the Eigen Vectors. Eigen Vectors show the direction of the MDLP component (SC Asian, S Asian, etc.) axes on the bi-plot, as well as the weights assigned to them by the software.

http://i1377.photobucket.com/albums/ah42/dil88888/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg (http://s1377.photobucket.com/user/dil88888/media/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg.html)

You will note that all MDLP components are well represented by PC1 & PC2, except for the E Eurasian & African comps. Thus, any member (Viktor) or ancient genomes [UST, Yamnaya (I included Altaic with N Asian),etc.] with high percentages of those components are not as well represented. PC3 & PC4 which I may post later, will give a better idea on placement of those genomes.

The Eigen Vectors which represent the various MDLP components. The representation of the MDLP components (SC Asia, S Asia, etc.) in each principal component is proportional to the length of the vector. The angle between the vectors represents the correlation of the components (S Asian, SC Asian, etc), with angles close to 0 representing that the 2 components/variables are positively correlated. In other words, an increase in one component is usually accompanied with an increase in the other.

I will be posting additional charts over the weekend. To further aid legibility, I have deactivated member names on some of the charts. So if you don't see your name on one chart, check the other charts.


http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg .html)

Kurd
03-21-2015, 11:04 AM
First, a little on Principal Component Analysis (PCA), also known as multi-variate analysis for those not familiar with it. Untill now, I have used 2D charts to plot members' admixture variations. 2D charts work well as long as the number of components is low. Since I have been using low k calculators so far, this has worked well, because I have been able to capture a high percentage of the members' admixture with a couple of charts, with each chart capturing 2 components.

With higher k calculators, I need to be able to reduce multiple dimensions (dimensionality of multiple components) down to a few dimensions or components. PCA does this by reducing the dimensionality of of multiple variables down to a few principal components. The PCA algorithm minimizes the loss of information/variability when the dimensionality of the calculator is reduced.

With the MDLP k23, I have been able to capture about 62% of the member's genetic variation with 2 prinicipal components. With the 1st 4 principal components, I am at around 90%. I may post a bi-plot showing PC3 and PC4 later. The software has weighted the components as follows:

http://i1377.photobucket.com/albums/ah42/dil88888/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg (http://s1377.photobucket.com/user/dil88888/media/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg.html)

For legibility reasons, I have de-activated the Eigen vector layer from the PCA bi-plot. For those interested, the following are the Eigen Vectors. Eigen Vectors show the direction of the MDLP component (SC Asian, S Asian, etc.) axes on the bi-plot, as well as the weights assigned to them by the software.

http://i1377.photobucket.com/albums/ah42/dil88888/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg (http://s1377.photobucket.com/user/dil88888/media/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg.html)

You will note that all MDLP components are well represented by PC1 & PC2, except for the E Eurasian & African comps. Thus, any member (Viktor) or ancient genomes [UST, Yamnaya (Altaic was included in N. Asian), etc.] with high percentages of those components are not as well represented. PC3 & PC4 which I may post later, will give a better idea on placement of those genomes.

The Eigen Vectors which represent the various MDLP components. The representation of the MDLP components (SC Asia, S Asia, etc.) in each principal component is proportional to the length of the vector. The angle between the vectors represents the correlation of the components (S Asian, SC Asian, etc), with angles close to 0 representing that the 2 components/variables are positively correlated. In other words, an increase in one component is usually accompanied with an increase in the other.

I will be posting additional charts over the weekend. To further aid legibility, I have deactivated member names on some of the charts. So if you don't see your name on one chart, check the other charts.


http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg .html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot%20Assyrian%20Turk%20Med%20C% 20Asian_zpsawzhvnym.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot%20Assyrian%20Turk%20Med%20C% 20Asian_zpsawzhvnym.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot%20European%20Armenian_zpsjn1 wntba.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot%20European%20Armenian_zpsjn1 wntba.jpg.html)

Kurd
03-21-2015, 11:29 AM
First, a little on Principal Component Analysis (PCA), also known as multi-variate analysis for those not familiar with it. Untill now, I have used 2D charts to plot members' admixture variations. 2D charts work well as long as the number of components is low. Since I have been using low k calculators so far, this has worked well, because I have been able to capture a high percentage of the members' admixture with a couple of charts, with each chart capturing 2 components.

With higher k calculators, I need to be able to reduce multiple dimensions (dimensionality of multiple components) down to a few dimensions or components. PCA does this by reducing the dimensionality of of multiple variables down to a few principal components. The PCA algorithm minimizes the loss of information/variability when the dimensionality of the calculator is reduced.

With the MDLP k23, I have been able to capture about 62% of the member's genetic variation with 2 prinicipal components. With the 1st 4 principal components, I am at around 90%. I may post a bi-plot showing PC3 and PC4 later. The software has weighted the components as follows:

http://i1377.photobucket.com/albums/ah42/dil88888/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg (http://s1377.photobucket.com/user/dil88888/media/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg.html)

For legibility reasons, I have de-activated the Eigen vector layer from the PCA bi-plot. For those interested, the following are the Eigen Vectors. Eigen Vectors show the direction of the MDLP component (SC Asian, S Asian, etc.) axes on the bi-plot, as well as the weights assigned to them by the software.

http://i1377.photobucket.com/albums/ah42/dil88888/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg (http://s1377.photobucket.com/user/dil88888/media/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg.html)

You will note that all MDLP components are well represented by PC1 & PC2, except for the E Eurasian & African comps. Thus, any member (Viktor) or ancient genomes [UST, Yamnaya (I included Altaic with N Asian),etc.] with high percentages of those components are not as well represented. PC3 & PC4 which I may post later, will give a better idea on placement of those genomes.

The Eigen Vectors which represent the various MDLP components. The representation of the MDLP components (SC Asia, S Asia, etc.) in each principal component is proportional to the length of the vector. The angle between the vectors represents the correlation of the components (S Asian, SC Asian, etc), with angles close to 0 representing that the 2 components/variables are positively correlated. In other words, an increase in one component is usually accompanied with an increase in the other.

I will be posting additional charts over the weekend. To further aid legibility, I have deactivated member names on some of the charts. So if you don't see your name on one chart, check the other charts.


http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg .html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot%20Pashtuns_zps7xljfgij.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot%20Pashtuns_zps7xljfgij.jpg.h tml)

Kurd
03-21-2015, 12:24 PM
First, a little on Principal Component Analysis (PCA), also known as multi-variate analysis for those not familiar with it. Untill now, I have used 2D charts to plot members' admixture variations. 2D charts work well as long as the number of components is low. Since I have been using low k calculators so far, this has worked well, because I have been able to capture a high percentage of the members' admixture with a couple of charts, with each chart capturing 2 components.

With higher k calculators, I need to be able to reduce multiple dimensions (dimensionality of multiple components) down to a few dimensions or components. PCA does this by reducing the dimensionality of of multiple variables down to a few principal components. The PCA algorithm minimizes the loss of information/variability when the dimensionality of the calculator is reduced.

With the MDLP k23, I have been able to capture about 62% of the member's genetic variation with 2 prinicipal components. With the 1st 4 principal components, I am at around 90%. I may post a bi-plot showing PC3 and PC4 later. The software has weighted the components as follows:

http://i1377.photobucket.com/albums/ah42/dil88888/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg (http://s1377.photobucket.com/user/dil88888/media/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg.html)

For legibility reasons, I have de-activated the Eigen vector layer from the PCA bi-plot. For those interested, the following are the Eigen Vectors. Eigen Vectors show the direction of the MDLP component (SC Asian, S Asian, etc.) axes on the bi-plot, as well as the weights assigned to them by the software.

http://i1377.photobucket.com/albums/ah42/dil88888/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg (http://s1377.photobucket.com/user/dil88888/media/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg.html)

You will note that all MDLP components are well represented by PC1 & PC2, except for the E Eurasian & African comps. Thus, any member (Viktor) or ancient genomes [UST, Yamnaya (I included Altaic with N Asian),etc.] with high percentages of those components are not as well represented. PC3 & PC4 which I may post later, will give a better idea on placement of those genomes.

The Eigen Vectors which represent the various MDLP components. The representation of the MDLP components (SC Asia, S Asia, etc.) in each principal component is proportional to the length of the vector. The angle between the vectors represents the correlation of the components (S Asian, SC Asian, etc), with angles close to 0 representing that the 2 components/variables are positively correlated. In other words, an increase in one component is usually accompanied with an increase in the other.

I will be posting additional charts over the weekend. To further aid legibility, I have deactivated member names on some of the charts. So if you don't see your name on one chart, check the other charts.


http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg .html)

I am also experimenting with 3D plots. Here is a preview. More to come later.

http://i1377.photobucket.com/albums/ah42/dil88888/3%20D%20S%20SC%20Asian%20N%20NE%20Asian_zps69okuoh g.jpg (http://s1377.photobucket.com/user/dil88888/media/3%20D%20S%20SC%20Asian%20N%20NE%20Asian_zps69okuoh g.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/3%20D%20S%20Asian%20SC%20Asian%20Caucasian_zpspxfs ch9m.jpg (http://s1377.photobucket.com/user/dil88888/media/3%20D%20S%20Asian%20SC%20Asian%20Caucasian_zpspxfs ch9m.jpg.html)

Kurd
03-21-2015, 04:40 PM
First, a little on Principal Component Analysis (PCA), also known as multi-variate analysis for those not familiar with it. Untill now, I have used 2D charts to plot members' admixture variations. 2D charts work well as long as the number of components is low. Since I have been using low k calculators so far, this has worked well, because I have been able to capture a high percentage of the members' admixture with a couple of charts, with each chart capturing 2 components.

With higher k calculators, I need to be able to reduce multiple dimensions (dimensionality of multiple components) down to a few dimensions or components. PCA does this by reducing the dimensionality of of multiple variables down to a few principal components. The PCA algorithm minimizes the loss of information/variability when the dimensionality of the calculator is reduced.

With the MDLP k23, I have been able to capture about 62% of the member's genetic variation with 2 prinicipal components. With the 1st 4 principal components, I am at around 90%. I may post a bi-plot showing PC3 and PC4 later. The software has weighted the components as follows:

http://i1377.photobucket.com/albums/ah42/dil88888/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg (http://s1377.photobucket.com/user/dil88888/media/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg.html)

For legibility reasons, I have de-activated the Eigen vector layer from the PCA bi-plot. For those interested, the following are the Eigen Vectors. Eigen Vectors show the direction of the MDLP component (SC Asian, S Asian, etc.) axes on the bi-plot, as well as the weights assigned to them by the software.

http://i1377.photobucket.com/albums/ah42/dil88888/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg (http://s1377.photobucket.com/user/dil88888/media/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg.html)

You will note that all MDLP components are well represented by PC1 & PC2, except for the E Eurasian & African comps. Thus, any member (Viktor) or ancient genomes [UST, Yamnaya (I included Altaic with N Asian),etc.] with high percentages of those components are not as well represented. PC3 & PC4 which I may post later, will give a better idea on placement of those genomes.

The Eigen Vectors which represent the various MDLP components. The representation of the MDLP components (SC Asia, S Asia, etc.) in each principal component is proportional to the length of the vector. The angle between the vectors represents the correlation of the components (S Asian, SC Asian, etc), with angles close to 0 representing that the 2 components/variables are positively correlated. In other words, an increase in one component is usually accompanied with an increase in the other.

I will be posting additional charts over the weekend. To further aid legibility, I have deactivated member names on some of the charts. So if you don't see your name on one chart, check the other charts.


http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg .html)

This 3 D graph captures more of Ust Ishim's admixture, and is thus better suited for comparison. If your name is not on the chart, it will be included on subsequent ones.

http://i1377.photobucket.com/albums/ah42/dil88888/3%20D%20S%20SC%20E%20Asian_zps3vsjj02r.jpg (http://s1377.photobucket.com/user/dil88888/media/3%20D%20S%20SC%20E%20Asian_zps3vsjj02r.jpg.html)

EDIT: I would like to caution members from inferring distance to Ust Ishim solely based on this 3D plot. Although, this plot captures a majority of his admixture, it does not capture all of it, as it is limited to 3 dimensions. E Asian includes all N E SE Asian comps. To get a more accurate distance to Ust, the distance from this plot should be added to the distance from a subsequent plot that captures his remaining components.

Kurd
03-21-2015, 06:21 PM
First, a little on Principal Component Analysis (PCA), also known as multi-variate analysis for those not familiar with it. Untill now, I have used 2D charts to plot members' admixture variations. 2D charts work well as long as the number of components is low. Since I have been using low k calculators so far, this has worked well, because I have been able to capture a high percentage of the members' admixture with a couple of charts, with each chart capturing 2 components.

With higher k calculators, I need to be able to reduce multiple dimensions (dimensionality of multiple components) down to a few dimensions or components. PCA does this by reducing the dimensionality of of multiple variables down to a few principal components. The PCA algorithm minimizes the loss of information/variability when the dimensionality of the calculator is reduced.

With the MDLP k23, I have been able to capture about 62% of the member's genetic variation with 2 prinicipal components. With the 1st 4 principal components, I am at around 90%. I may post a bi-plot showing PC3 and PC4 later. The software has weighted the components as follows:

http://i1377.photobucket.com/albums/ah42/dil88888/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg (http://s1377.photobucket.com/user/dil88888/media/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg.html)

For legibility reasons, I have de-activated the Eigen vector layer from the PCA bi-plot. For those interested, the following are the Eigen Vectors. Eigen Vectors show the direction of the MDLP component (SC Asian, S Asian, etc.) axes on the bi-plot, as well as the weights assigned to them by the software.

http://i1377.photobucket.com/albums/ah42/dil88888/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg (http://s1377.photobucket.com/user/dil88888/media/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg.html)

You will note that all MDLP components are well represented by PC1 & PC2, except for the E Eurasian & African comps. Thus, any member (Viktor) or ancient genomes [UST, Yamnaya (I included Altaic with N Asian),etc.] with high percentages of those components are not as well represented. PC3 & PC4 which I may post later, will give a better idea on placement of those genomes.

The Eigen Vectors which represent the various MDLP components. The representation of the MDLP components (SC Asia, S Asia, etc.) in each principal component is proportional to the length of the vector. The angle between the vectors represents the correlation of the components (S Asian, SC Asian, etc), with angles close to 0 representing that the 2 components/variables are positively correlated. In other words, an increase in one component is usually accompanied with an increase in the other.

I will be posting additional charts over the weekend. To further aid legibility, I have deactivated member names on some of the charts. So if you don't see your name on one chart, check the other charts.


http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot_zpswkvxgsgj.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg .html)

With the addition of principal components 3 and 4 to PC 1 and PC2, I am able to cover around 90% of the members' admixture variation. Distance to Ust-Ishim or to any other member can reasonably be inferred by adding the distances from the previous PCA (PC1 & PC2) to the distance in this plot (PC3 & PC4). PC 3 and PC 4 capture more of the E Eurasian and African components.

http://i1377.photobucket.com/albums/ah42/dil88888/PC3%20PC4_zpseopyd3tr.jpg (http://s1377.photobucket.com/user/dil88888/media/PC3%20PC4_zpseopyd3tr.jpg.html)

Edit:The approximate x co-ordinate of Ust Ishim is 9 (he is way out there on the x axis)

Kurd
03-22-2015, 02:04 AM
First, a little on Principal Component Analysis (PCA), also known as multi-variate analysis for those not familiar with it. Untill now, I have used 2D charts to plot members' admixture variations. 2D charts work well as long as the number of components is low. Since I have been using low k calculators so far, this has worked well, because I have been able to capture a high percentage of the members' admixture with a couple of charts, with each chart capturing 2 components.

With higher k calculators, I need to be able to reduce multiple dimensions (dimensionality of multiple components) down to a few dimensions or components. PCA does this by reducing the dimensionality of of multiple variables down to a few principal components. The PCA algorithm minimizes the loss of information/variability when the dimensionality of the calculator is reduced.

With the MDLP k23, I have been able to capture about 62% of the member's genetic variation with 2 prinicipal components. With the 1st 4 principal components, I am at around 90%. I may post a bi-plot showing PC3 and PC4 later. The software has weighted the components as follows:

http://i1377.photobucket.com/albums/ah42/dil88888/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg (http://s1377.photobucket.com/user/dil88888/media/Eigenvalues%20amp%20PC%20Coefficients_zpshzb4xx0x. jpg.html)

For legibility reasons, I have de-activated the Eigen vector layer from the PCA bi-plot. For those interested, the following are the Eigen Vectors. Eigen Vectors show the direction of the MDLP component (SC Asian, S Asian, etc.) axes on the bi-plot, as well as the weights assigned to them by the software.

http://i1377.photobucket.com/albums/ah42/dil88888/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg (http://s1377.photobucket.com/user/dil88888/media/Eigen%20vectors%20PC1%20PC2_zps5i2kc6sp.jpg.html)

You will note that all MDLP components are well represented by PC1 & PC2, except for the E Eurasian & African comps. Thus, any member (Viktor) or ancient genomes [UST, Yamnaya (I included Altaic with N Asian),etc.] with high percentages of those components are not as well represented. PC3 & PC4 which I may post later, will give a better idea on placement of those genomes.

The representation of the MDLP components (SC Asia, S Asia, etc.) in each principal component is proportional to the length of the vector. The angle between the vectors represents the correlation of the components (S Asian, SC Asian, etc), with angles close to 0 representing that the 2 components/variables are positively correlated. In other words, an increase in one component is usually accompanied with an increase in the other.

I will be posting additional charts over the weekend. To further aid legibility, I have deactivated member names on some of the charts. So if you don't see your name on one chart, check the other charts.


http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20Bi%20Plot%20S%20Asians_zps87zh80gl.jpg .html)

The following shows the Eigenvectors for PC3 & PC4. SE Asian and African are positively correlated (small angle), and well represented (large vector sizes) by principal components 3 & 4. N/NE Asian is also well represented here, but not correlated with the former two (90 deg angle). The x axis is heavy with SE Asian & African, and the y axis is heavy with N /NE Asian.

The remaining components of MDLP are better represented by PC1 & PC2. Between the 2 graphs, the program captured about 90% of the members' admixture variation. Thus, the two graphs taken together can be used to infer distance to other members and ancient genomes. It is important that you add the distance using both graphs.

To illustrate, we can take a look at Rukha's & Armenian 1's distance to Yamnaya. The first graph (PC3/PC4) shows both to be relatively close to Yamnaya. However, if we take a look at the second graph (PC1/PC2), to add the distances, we can see that Rukha has a much closer total distance to Yamnaya than Armenian1. This is consistent with Davidski's results for Rukha, showing him to be relatively close to Yamnaya.

Also, for those wishing to check their distance from Ust Ishim. His x co-ordinate is around 9.

http://i1377.photobucket.com/albums/ah42/dil88888/Eigenvectors%20PC3%20PC4_zpsb829wkkj.jpg (http://s1377.photobucket.com/user/dil88888/media/Eigenvectors%20PC3%20PC4_zpsb829wkkj.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC3%20PC4%20Europeans_zpsphwwwvlm.jpg (http://s1377.photobucket.com/user/dil88888/media/PC3%20PC4%20Europeans_zpsphwwwvlm.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20YAMNAYA_zpszxm7dpvx.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20YAMNAYA_zpszxm7dpvx.jpg.html)

Kurd
03-22-2015, 04:25 AM
Here are a few MDLP k23 community admixture graphs with a few ancient genomes:

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20%20Community%20S%20E%20N%20ASIAN%20AD MIX_zpsztvymyj2.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20%20Community%20S%20E%20N%20ASIAN%20AD MIX_zpsztvymyj2.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20%20Community%20African%20ADMIX_zpscnq 7zmwy.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20%20Community%20African%20ADMIX_zpscnq 7zmwy.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20%20Community%20SW%20ASIAN%20ADMIX_zps jtvknxh5.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20%20Community%20SW%20ASIAN%20ADMIX_zps jtvknxh5.jpg.html)

parasar
03-22-2015, 06:00 AM
...

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20%20Community%20S%20E%20N%20ASIAN%20AD MIX_zpsztvymyj2.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20%20Community%20S%20E%20N%20ASIAN%20AD MIX_zpsztvymyj2.jpg.html)

...[/IMG]

Ust-Ishim's proportions look like that of an Austro-Asiatic from eastern India.

Quite in line with:
http://eurogenes.blogspot.com/2014/11/treemix-graphs-with-kostenki14-and-ust.html


Epicenter is somewhere in Burma. Interestingly, Papuans and Melanesians are close.
Munda-South 57 ---//Confidence: medium
Lahu 56,66 ---//Confidence: low
Malayan 55,72 ---//Confidence: very low
Nivh 55,08 ---//Confidence: very low
Munda-North 54,39 ---//Confidence: low
Hazara 53,34 ---//Confidence: very high
Burmese 53,33 ---//Confidence: high


http://s020.radikal.ru/i700/1411/99/39e84d81c245.png

I also think that the African affinity in Ust-Ishim is real. I think this is old African affinity rather than that of the recent type seen in the Makran to Horn area.
How close are the Hazara to Ust-Ishim?

Helgenes50
03-22-2015, 07:03 AM
Thanks for your work that is very interesting

I know I am late, but here are my results



# Population Percent

1 European_Hunters_Gatherers 34.29
2 European_Early_Farmers 31.87
3 Caucasian 22.44
4 Ancestral_Altaic 4.26
5 South_Central_Asian 2.88
6 Near_East 1.32
7 North_African 1.23
8 Amerindian 1.12



--------------------------------

Least-squares method.

Using 1 population approximation:
1 English_Kent_GBR @ 4.083309
2 English_Cornwall_GBR @ 5.092705
3 Belgian @ 5.138433
4 Welsh @ 5.700315
5 English @ 5.855653
6 Frisian @ 6.137664
7 CEU @ 6.157896
8 British @ 6.238829
9 North_European @ 6.590234
10 Irish @ 6.649572
11 French @ 7.042384
12 South_German @ 7.585067
13 German-Volga @ 7.699992
14 Scottish_Argyll_Bute_GBR @ 8.894088
15 Orcadian @ 8.895553
16 Dutch @ 9.428334
17 Norwegian_West @ 10.737793
18 Norwegian_East @ 11.169126
19 Icelandic @ 11.561574
20 North_German @ 12.636159

Using 2 populations approximation:
1 50% French +50% South_German @ 2.493501

Kurd
03-22-2015, 12:19 PM
Ust-Ishim's proportions look like that of an Austro-Asiatic from eastern India.

Quite in line with:
http://eurogenes.blogspot.com/2014/11/treemix-graphs-with-kostenki14-and-ust.html

http://s020.radikal.ru/i700/1411/99/39e84d81c245.png

I also think that the African affinity in Ust-Ishim is real. I think this is old African affinity rather than that of the recent type seen in the Makran to Horn area.
How close are the Hazara to Ust-Ishim?

The problem is that due to Ust Ishim's age, it is hard to find a modern population with his rather odd combination (by modern standards) of ancestral alleles. His MDLP k23 results were:

South_Indian 28.37
Australoid 8.74
South_East_Asian 7.00
European_Early_Farmers 6.64
Austronesian 6.54
European_Hunters_Gatherers 5.94
East_African 4.83
South_Central_Asian 4.78
Melano_Polynesian 4.61
Archaic_African 4.14
Near_East 3.02
Subsaharian 2.93
North_African 2.57
Paleo_Siberian 2.34
Tungus-Altaic 2.04
Arctic 1.44
Ancestral_Altaic 1.28
Archaic_Human 1.12
Amerindian 1.11

His proxies were:

1 Ayta_AE @ 24.783888
2 Nepalese @ 27.270012
3 Kensiu @ 29.925926
4 Tamil_Singapore @ 29.993927
5 Cochin_Jew @ 30.449663
6 Onge @ 31.218340
7 Pahari @ 31.865376

With this calculator, he has lots of African & S Indian, with substantial Australoid, Austranesian, SE Asian, and EEF. The sole Hazara individual that was part of this project showed an overlap of about 28% with him, which is less than my 40%, and your 47%.

For my next project, I will target a calculator rich in SE Asian and S Asian references, which will probably assist in this regard, as well as give us an additional set of results for our comparison.

Kurd
03-22-2015, 02:14 PM
Here are graphs showing the community admixture overlap/similarity to SC Asians and Iranians/Kurds. This time around I included Viktor and the Hazara individual in the SC Asian reference. Thus, the SC Asian reference is the average of all the Pashtun project members, Viktor, and Hazara. This addition added some E Eurasian into the mean. Noteworthy, is the substantial overlap of Yamnaya with this reference.


http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20%20Community%20Similarity%20SC%20Asia ns_zps68cryrzx.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20%20Community%20Similarity%20SC%20Asia ns_zps68cryrzx.jpg.html)

The Iranian / Kurd reference included all the Iranian & Kurd project members, excluding my daughter. Noteworthy, is the number of ancients with a decent admixture overlap with this reference.


http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20%20Community%20Similarity%20Iranians_ zpstzunrrlw.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20%20Community%20Similarity%20Iranians_ zpstzunrrlw.jpg.html)

EDIT: Corrected Iranian/Kurd graph
2nd edit: The Iranian Baloch individuals that were included in the reference pushed me and Jesus close to the Iranian/Kurd mean, which is shown as high scores for us on the graph

Kurd
03-22-2015, 03:50 PM
Here are graphs showing the community admixture overlap/similarity to SC Asians and Iranians/Kurds. This time around I included Viktor and the Hazara individual in the SC Asian reference. Thus, the SC Asian reference is the average of all the Pashtun project members, Viktor, and Hazara. This addition added some E Eurasian into the mean. Noteworthy, is the substantial overlap of Yamnaya with this reference.


http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20%20Community%20Similarity%20SC%20Asia ns_zps68cryrzx.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20%20Community%20Similarity%20SC%20Asia ns_zps68cryrzx.jpg.html)

The Iranian / Kurd reference included all the Iranian & Kurd project members, excluding my daughter. Noteworthy, is the number of ancients with a decent admixture overlap with this reference.


http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20%20Community%20Similarity%20Iranians_ zpstzunrrlw.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20%20Community%20Similarity%20Iranians_ zpstzunrrlw.jpg.html)

EDIT: Corrected Iranian/Kurd graph
2nd edit: The Iranian Baloch individuals that were included in the reference pushed me and Jesus close to the Iranian/Kurd mean, which is shown as high scores for us on the graph

Here is a graph comparing the community's admixture to S Asians. For the S Asian reference, I used all the project members, except Pashtuns with Indian and Pakistani ancestry.

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20%20Community%20Admix%20Similarity%20t o%20S%20Asians_zpsnk8qcjdn.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20%20Community%20Admix%20Similarity%20t o%20S%20Asians_zpsnk8qcjdn.jpg.html)

The Barnacle
03-22-2015, 04:58 PM
Kurd what would happen if you don't include Sein and Adam to the South Central ashan reference?

Kurd
03-22-2015, 05:21 PM
Kurd what would happen if you don't include Sein and Adam to the South Central ashan reference?

A couple of things, since the reference sample is relatively small:

1- Sein & Adam's SC Asian score will decrease. The same may occur for S Asians;
2- SC Asian score for Iranians and others will increase.

Edit: If the reference sample would have been much larger, then their removal would have a much smaller impact.

vettor
03-22-2015, 05:42 PM
@kurd

do you have the mix for NE1_7KY ...................I see to have very similar to it

Arbogan
03-22-2015, 05:43 PM
id like to participate if possible.- edit just saw that you need to post data.

Population:
Ancestral_Altaic 4.09%
South_Central_Asian 27.52%
Arctic 0.69%
South_Indian 4.42%
Caucasian 38.14%
East_Siberian 0.52%
European_Early_Farmers 4.88%
Archaic_African 0.18%
Near_East 12.79%
North_African 2.05%
Paleo_Siberian 0.36%
African_Pygmy 0.42%
South_East_Asian 0.41%
Subsaharian 1.07%
Tungus-Altaic 0.64%
European_Hunters_Gatherers 1.78%

Dr_McNinja
03-22-2015, 07:29 PM
Here are a few MDLP k23 community admixture graphs with a few ancient genomes:

http://i1377.photobucket.com/albums/ah42/dil88888/MDLP%20K23%20%20Community%20S%20E%20N%20ASIAN%20AD MIX_zpsztvymyj2.jpg (http://s1377.photobucket.com/user/dil88888/media/MDLP%20K23%20%20Community%20S%20E%20N%20ASIAN%20AD MIX_zpsztvymyj2.jpg.html)


Did N/NE Asian here include Ancestral Altaic/ANE?

Kurd
03-22-2015, 08:12 PM
Did N/NE Asian here include Ancestral Altaic/ANE?

no, it did'nt

The Barnacle
03-22-2015, 08:51 PM
A couple of things, since the reference sample is relatively small:

1- Sein & Adam's SC Asian score will decrease. The same may occur for S Asians;
2- SC Asian score for Iranians and others will increase.

Edit: If the reference sample would have been much larger, then their removal would have a much smaller impact.

I know this is debateable but doesn't South Central Asia refer to only afghanistan? Pakistan South Asia?

ViktorL1
03-22-2015, 09:01 PM
It seems I am quite... the outlier. I'm surprised that my East Asian components are substantially greater than Hazara's. Seems to contradict the historical/social narrative.

4121
4122

Dr_McNinja
03-22-2015, 09:23 PM
I know this is debateable but doesn't South Central Asia refer to only afghanistan? Pakistan South Asia?
South Central Asia isn't really an "official" term like South Asia or Central Asia and usually refers to everything between Central Asia and Balochistan pretty much. Gedrosian/Baloch in admixture calculators like Eurogenes K36 (along with all the HGDP Pakistan populations) is labeled "South Central" Asian.

It's basically that part of South and Central Asia (http://www.state.gov/p/sca/ci/) which is between Central Asia (http://www.britannica.com/EBchecked/topic/102288/Central-Asia) and the Indian subcontinent (since this entire area is usually labeled South Asia (https://www.cia.gov/library/publications/the-world-factbook/wfbExt/region_sas.html), but there are two components in this region, so the one to the west (usually Balochistan) is termed South Central Asian as the Indian one gets "South Asian").

Kurd
03-22-2015, 09:23 PM
I know this is debateable but doesn't South Central Asia refer to only afghanistan? Pakistan South Asia?

You are absolutely correct. Pakistan is within S. Asia. Sein and Adam's S Indian scores of 25% are closer to the Pakistani individuals Sindhi, Raskol, and even a little closer to Pak Jatt, Sapporo, and Hary Jatt, than to Zahra and Kandhari. I actually have included them in the S Asian reference previously. I included them with the SC Asian reference this time because I only had a couple of Afghan Pashtun members, and several Indian and Pakistani members.

This will not be an issue when I base my projects on calculators that have rich SC and S Asian references. If I have time, I will redo the chart, and include them in the S Asian reference. This will increase SC Asian scores & S Asian scores for most. The effect will be more pronounced on SC Asian scores due to the small reference sample.

Kurd
03-22-2015, 09:26 PM
It seems I am quite... the outlier. I'm surprised that my East Asian components are substantially greater than Hazara's. Seems to contradict the historical/social narrative.

4121
4122

Your E Asian components as well as your S Indian ;). I am not familiar with this individual. His data was supplied by Rukha. He may be familiar with his background.

ViktorL1
03-22-2015, 09:46 PM
Your E Asian components as well as your S Indian ;). I am not familiar with this individual. His data was supplied by Rukha. He may be familiar with his background.

As in that component is elevated or depressed? I have no reference for other Southern Central Asians (by that I mean Fergana Uzbeks/Tajiks).

Dr_McNinja
03-22-2015, 09:52 PM
Kurd, could you make a bar graph going to 100% to show admixture "layers" like the ones showing various K? You could group the east Asian/African/Arctic together to minimize components but I'd suggest keeping Caucasian, Gedrosian, and Indian separate to show the difference in the group better (also Near East + North African are basically both "SW-Asian" and quite Caucasus like).

Rukha
03-22-2015, 09:54 PM
Your E Asian components as well as your S Indian ;). I am not familiar with this individual. His data was supplied by Rukha. He may be familiar with his background.

He's an Hazara from Ghazni which is one of the more diverse provinces in Afghanistan. He is R1a1a* paternal and M2a1a maternal.

edit: I presented his results in the wrong order Kurd,his East Eurasian score is higher than what that chart suggests


Population
Amerindian 0.32%
Ancestral_Altaic 4.74%
South_Central_Asian 17.36%
Arctic 2.05%
South_Indian 5.89%
Australoid 0.87%
Austronesian 3.00%
Caucasian 12.90%
Archaic_Human -
East_African -
East_Siberian 7.25%
European_Early_Farmers 2.46%
Khoisan -
Melano_Polynesian -
Archaic_African 0.06%
Near_East 1.20%
North_African -
Paleo_Siberian 0.96%
African_Pygmy -
South_East_Asian 14.21%
Subsaharian -
Tungus-Altaic 20.25%
European_Hunters_Gatherers 6.47%

Kurd
03-22-2015, 10:03 PM
Kurd, could you make a bar graph going to 100% to show admixture "layers" like the ones showing various K? You could group the east Asian/African/Arctic together to minimize components but I'd suggest keeping Caucasian, Gedrosian, and Indian separate to show the difference in the group better (also Near East + North African are basically both "SW-Asian" and quite Caucasus like).

sure no problem I'll do that when I get to my computer later today

parasar
03-22-2015, 11:26 PM
The problem is that due to Ust Ishim's age, it is hard to find a modern population with his rather odd combination (by modern standards) of ancestral alleles. His MDLP k23 results were:

South_Indian 28.37
Australoid 8.74
South_East_Asian 7.00
European_Early_Farmers 6.64
Austronesian 6.54
European_Hunters_Gatherers 5.94
East_African 4.83
South_Central_Asian 4.78
Melano_Polynesian 4.61
Archaic_African 4.14
Near_East 3.02
Subsaharian 2.93
North_African 2.57
Paleo_Siberian 2.34
Tungus-Altaic 2.04
Arctic 1.44
Ancestral_Altaic 1.28
Archaic_Human 1.12
Amerindian 1.11

Yes I agree Ust-Ishim is quite different.

Modern populations tend to form their own components. For example Native Americans and South Asians share a lot of ancestry but when seen in terms of modern components are quite distinct, especially the Amerindians who underwent a severe bottleneck.

But when we compare these populations in terms of ANE, ENF, WHG, SSA then the shared ancestry (or lack thereof) becomes apparent. Interestingly SSA starts showing up in South Asians and that too at higher levels in Austric South Asians who are more distant geographically from Africa than western South Asians.




With this calculator, he has lots of African & S Indian, with substantial Australoid, Austranesian, SE Asian, and EEF. The sole Hazara individual that was part of this project showed an overlap of about 28% with him, which is less than my 40%, and your 47% ...

Regarding the Hazara what surprised me is their affinity to Ust-Ishim for >3cM segments - much higher than their neighbors have. One reason may be that the Hazara with a recent inner-east Asiatic influence are indeed quite distinct from their neighbors. Though I have a nagging suspicion that the Hazara harbor some old lines - their Y-B-M60 and Y-M1 may also have something to do with it.

seferhabahir
03-23-2015, 04:55 AM
@kurd

do you have the mix for NE1_7KY ...................I see to have very similar to it

Go here...

http://www.fi.id.au/2014/11/ne1-ancient-dna-analysis.html

The Barnacle
03-23-2015, 08:15 AM
You are absolutely correct. Pakistan is within S. Asia. Sein and Adam's S Indian scores of 25% are closer to the Pakistani individuals Sindhi, Raskol, and even a little closer to Pak Jatt, Sapporo, and Hary Jatt, than to Zahra and Kandhari. I actually have included them in the S Asian reference previously. I included them with the SC Asian reference this time because I only had a couple of Afghan Pashtun members, and several Indian and Pakistani members.

This will not be an issue when I base my projects on calculators that have rich SC and S Asian references. If I have time, I will redo the chart, and include them in the S Asian reference. This will increase SC Asian scores & S Asian scores for most. The effect will be more pronounced on SC Asian scores due to the small reference sample.

That would make more sense IMO thank you Kurd

Kurd
03-23-2015, 05:41 PM
Here is a graph as requested by Dr McNinja showing community admixture. W Asian is defined as Near East & Caucasus. European as EHG & EEF. E Asian includes all asian comps except Anc Altaic (not represented in graph). Baibars & 2 Kurds, Feyli 1&2, have been added. Hazara has been corrected.

http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20ADMIX%20ASIAN%20EUROPEAN_zpsotdctuuk.j pg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20ADMIX%20ASIAN%20EUROPEAN_zpsotdctuuk.j pg.html)

The following has the following revisions; Hazara correction. Addition of Baibars, Feyli1 &2.


http://i1377.photobucket.com/albums/ah42/dil88888/COMMUNITY%20OVERLAP%20UST%20ISHIM_zpsmwkvdsrj.jpg (http://s1377.photobucket.com/user/dil88888/media/COMMUNITY%20OVERLAP%20UST%20ISHIM_zpsmwkvdsrj.jpg. html)

vettor
03-23-2015, 05:45 PM
I must be either Bactrian or Sogtrian with the amount of west-asian in me ............or should I say.... tajik/uzbek?

ZephyrousMandaru
03-23-2015, 11:06 PM
MDLP K23b


Population
Amerindian -
Ancestral_Altaic 2.11%
South_Central_Asian 18.08%
Arctic -
South_Indian 0.36%
Australoid 1.04%
Austronesian 0.06%
Caucasian 44.03%
Archaic_Human -
East_African -
East_Siberian -
European_Early_Farmers 11.42%
Khoisan 0.69%
Melano_Polynesian 0.18%
Archaic_African 0.30%
Near_East 17.19%
North_African 3.84%
Paleo_Siberian 0.65%
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers -

Kurd
03-24-2015, 05:03 AM
Here is a dendogram for the project participants. There are W Asian, C /SC Asian, S Asian, and European clusters. The program has assigned some Iranians/Kurds to the W Asian cluster and some to the C/SC Asian cluster based on distance.

For those not familiar with dendograms, I am giving a brief explanation. To find your distance to a neighboring member or population cluster, find the nearest horizontal "bridge" that connects you to the member or cluster, and read the distance on the y axis. To illustrate, I used myself as an example. My closest member is Jesus. My distance to him is about 0.08.


http://i1377.photobucket.com/albums/ah42/dil88888/Dendogram_zpswdjo7ptu.jpg (http://s1377.photobucket.com/user/dil88888/media/Dendogram_zpswdjo7ptu.jpg.html)

Humanist
03-24-2015, 05:22 AM
Here is a dendogram for the project participants. There are W Asian, C /SC Asian, S Asian, and European clusters. The program has assigned some Iranians/Kurds to the W Asian cluster and some to the C/SC Asian cluster based on distance.

Thanks. Interesting that the Iraqi Mandaeans and I branch off into a cluster that is more similar to the Iranians/Kurds, rather than to the Armenians/Turks.

Kurd
03-24-2015, 05:28 AM
Thanks. Interesting that the Iraqi Mandaeans and I branch off into a cluster that is more similar to the Iranians/Kurds, rather than to the Armenians/Turks.

Using the Eucladian method, which is what was used for the analysis, It looks like you are closer to the Iranian/Kurd average than the Armenian/Turk average comprised of those cluster members :)

Jusarius
03-24-2015, 06:26 AM
Just in case you decide to take in more data, here are my (East Finnish) results:

Population
Amerindian 0.54%
Ancestral_Altaic 7.79%
South_Central_Asian 0.66%
Arctic -
South_Indian 2.09%
Australoid 0.14%
Austronesian 0.84%
Caucasian 19.68%
Archaic_Human -
East_African -
East_Siberian 6.47%
European_Early_Farmers 8.51%
Khoisan -
Melano_Polynesian -
Archaic_African -
Near_East -
North_African 0.49%
Paleo_Siberian 0.77%
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic 0.41%
European_Hunters_Gatherers 51.59%

Using 1 population approximation:
1 Finnish-East @ 1.316280
2 Finn_East @ 1.347095
3 Finnish_FIN @ 3.688726
4 Finn @ 5.014487
5 Russian-Ural @ 6.179489
6 Russian-North @ 6.315299
7 Erzya @ 6.791250
8 Russian-Upper-Volga @ 7.466588
9 Finn_West @ 8.010735
10 Russian_Vologda @ 8.228518
11 Karelian @ 8.405813
12 Russian_Meshtchyora @ 9.217772
13 Saami_Kola @ 9.632266
14 Russian_North @ 9.838916
15 Russian-North-West @ 10.043378
16 Vepsa @ 10.146551
17 Russian-West @ 10.188738
18 Mordovian @ 10.853497
19 Moksha @ 11.075132
20 Belarusian-East @ 11.802301

The Barnacle
03-24-2015, 09:41 AM
Here is a dendogram for the project participants. There are W Asian, C /SC Asian, S Asian, and European clusters. The program has assigned some Iranians/Kurds to the W Asian cluster and some to the C/SC Asian cluster based on distance.

For those not familiar with dendograms, I am giving a brief explanation. To find your distance to a neighboring member or population cluster, find the nearest horizontal "bridge" that connects you to the member or cluster, and read the distance on the y axis. To illustrate, I used myself as an example. My closest member is Jesus. My distance to him is about 0.08. My distance to the average of the Pathan cluster would be ~0.27. To the average of the S Asian cluster/ Ust Ishim cluster is 0.58, and so on.


http://i1377.photobucket.com/albums/ah42/dil88888/Dendogram_zpswdjo7ptu.jpg (http://s1377.photobucket.com/user/dil88888/media/Dendogram_zpswdjo7ptu.jpg.html)

Am I the only one who doesn't understand these diagrams?

fil
03-24-2015, 12:27 PM
Am I the only one who doesn't understand these diagrams?

It took me awhile to figure out myself. It displays how similar you are to other members. So for example on mine, I am most similar to soulbrighter. Our chunk is most similar to varun and paraser and so on. If I didn't make much sense you can also check out this reference (http://wheatoncollege.edu/lexomics/files/2012/08/How-to-Read-a-Dendrogram-Web-Ready.pdf) which explains it quite nicely.

Speaking of which I am surprised that me and soulbrighter are linked together and not varun and soulbrighter because they were closer on the PCA and my lack of a european component.

Kurd
03-24-2015, 02:01 PM
[SIZE=3]Here is a dendogram for the project participants. There are W Asian, C /SC Asian, S Asian, and European clusters. The program has assigned some Iranians/Kurds to the W Asian cluster and some to the C/SC Asian cluster based on distance.

For those not familiar with dendograms, I am giving a brief explanation. To find your distance to a neighboring member or population cluster, find the nearest horizontal "bridge" that connects you to the member or cluster, and read the distance on the y axis. To illustrate, I used myself as an example. My closest member is Jesus. My distance to him is about 0.08.

http://i1377.photobucket.com/albums/ah42/dil88888/Dendogram_zpswdjo7ptu.jpg (http://s1377.photobucket.com/user/dil88888/media/Dendogram_zpswdjo7ptu.jpg.html)

There are several methods of hierchal cluster analysis. The above output was obtained using the Eucladian method. The output varies depending on the method. I will later be posting outputs using other methods such as cosine, pearson, etc.

DMXX
03-24-2015, 02:08 PM
Looks like the program has split the Iranians and Kurds based on the total South Asian-related scores.

Hanna
03-24-2015, 02:44 PM
http://i1377.photobucket.com/albums/ah42/dil88888/Dendogram_zpswdjo7ptu.jpg (http://s1377.photobucket.com/user/dil88888/media/Dendogram_zpswdjo7ptu.jpg.html)

So my closest match is my aunt. I would have expected some of the Armenians to match with Humanist. Like Armenian 6, Armenian 4 and Armenian 5 have higher southwest Asian and depending on the calculator match with Assyrians first.

Armenian5 k23b oracle 4:


# Population Percent
1 Caucasian 45.91
2 South_Central_Asian 19.09
3 Near_East 14.74
4 European_Early_Farmers 11.58
5 North_African 3.61
6 European_Hunters_Gatherers 1.25
7 South_Indian 1.10


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Assyrian_Iraqi @ 3.661483
2 Jew_Tat @ 5.746512
3 Iraqi_Chaldean @ 6.488720
4 Georgian_Jew @ 6.886033
5 Uzbekistani_Jew @ 7.278107
6 Turk_Kayseri @ 7.540285
7 Assyrian_Arzni @ 7.803115
8 Iranian_Jew @ 7.898488
9 Kurd_Jew @ 8.035158
10 Armenian @ 8.059820
11 Turk_Adana @ 8.075055
12 Iraqi_Jew @ 8.387489
13 Armenian_Yerevan @ 8.462649
14 Iraqi_Mandean @ 8.942708
15 Kurd_North @ 10.512911
16 Turk @ 10.762104
17 Lebanese_Druze @ 11.161404
18 Uzbek_Tashkent @ 11.335862
19 Azeri @ 11.541718
20 Lebanese_Muslim @ 11.648014

Using 2 populations approximation:
1 50% Armenian +50% Iraqi_Chaldean @ 3.384671


Using 3 populations approximation:
1 50% Assyrian_Arzni +25% Greek_Smyrna +25% Kurd_East @ 2.983840


Using 4 populations approximation:
1 Armenian + Assyrian_Iraqi + Cypriot + Kurd @ 2.245507
2 Armenian + Assyrian_Iraqi + Assyrian_Iraqi + Assyrian_Iraqi @ 2.410995
3 Armenian_Yerevan + Assyrian_Iraqi + Cypriot + Kurd @ 2.419417
4 Armenian + Assyrian_Iraqi + Cypriot + Kurd_East @ 2.439855
5 Armenian_Yerevan + Assyrian_Iraqi + Cypriot + Kurd_East @ 2.573558
6 Assyrian_Arzni + Assyrian_Iraqi + Cypriot + Kurd @ 2.616815
7 Armenian + Cypriot + Iraqi_Chaldean + Kurd @ 2.639545
8 Armenian + Assyrian_Iraqi + Assyrian_Iraqi + Turk_Kayseri @ 2.720357
9 Armenian + Assyrian_Iraqi + Assyrian_Iraqi + Turk_Adana @ 2.728396
10 Armenian + Assyrian_Iraqi + Cypriot + Kurd_North @ 2.749937
11 Armenian + Assyrian_Iraqi + Cypriot + Kurd_South @ 2.786530
12 Armenian_Yerevan + Assyrian_Iraqi + Cypriot + Kurd_North @ 2.795814
13 Assyrian_Arzni + Assyrian_Iraqi + Cypriot + Kurd_East @ 2.799362
14 Armenian + Assyrian_Iraqi + Assyrian_Iraqi + Iraqi_Chaldean @ 2.807998
15 Assyrian_Arzni + Assyrian_Iraqi + Greek_Smyrna + Kurd @ 2.825398
16 Armenian + Cypriot + Jew_Tat + Kurd_East @ 2.844767
17 Armenian + Cypriot + Jew_Tat + Kurd @ 2.847096
18 Armenian_Yerevan + Assyrian_Iraqi + Assyrian_Iraqi + Assyrian_Iraqi @ 2.864658
19 Armenian + Cypriot + Iraqi_Chaldean + Kurd_East @ 2.870969
20 Armenian + Cypriot + Iranian_Jew + Kurd @ 2.873236

seferhabahir
03-24-2015, 03:48 PM
Interesting that in the dendogram, I am now pulled over to group with the Turks and away from the other Mediterranean. This appears to be consistent with the earlier PCA1&2 graphs where I am also closer to the Turks and Armenians than to the other Mediterranean. Probably because of more West Asian I would assume.

Kurd
03-24-2015, 04:35 PM
Here is a dendogram for the project participants. There are W Asian, C /SC Asian, S Asian, and European clusters. The program has assigned some Iranians/Kurds to the W Asian cluster and some to the C/SC Asian cluster based on distance.

For those not familiar with dendograms, I am giving a brief explanation. To find your distance to a neighboring member or population cluster, find the nearest horizontal "bridge" that connects you to the member or cluster, and read the distance on the y axis. To illustrate, I used myself as an example. My closest member is Jesus. My distance to him is about 0.08.


http://i1377.photobucket.com/albums/ah42/dil88888/Dendogram_zpswdjo7ptu.jpg (http://s1377.photobucket.com/user/dil88888/media/Dendogram_zpswdjo7ptu.jpg.html)

Based on some of the responses I am seeing, I felt that I needed to further clarify the method used in this analysis.

First, the distance measure that was used to measure distance between members, is the Euclidean method, probably the most chosen method. It is basically the geometric distance in the multidimensional space (think of each of your admix components as a dimension).

Second, the linkage rule used for determining distance between clusters is the Complete linkage (furthest neighbor) method. In this method, the distances between clusters are determined by the greatest distance between any two objects in the different clusters (i.e., by the "furthest neighbors"). For example, the distance between the SC/S Asian and W Asian cluster was large because it was probably based on the distance between perhaps Fil or Ust Ishim and Armenian.

I will redo the analysis later using a different method of linkage analysis, maybe the Unweighted pair-group centroid method, where the centroid of a cluster is the average point in the multidimensional space defined by the dimensions. In a sense, it is the center of gravity for the respective cluster. In this method, the distance between two clusters is determined as the difference between centroids.

I will also post the analysis with and without the ancient genomes.

I will also be updating the PCA plots, using unreduced/consolidated admixture components. The accuracy should increase. the trade off may be that I will need 3 plots, PC1 - PC6 to capture a high percentage of the admixture variation. Thus, members would have to add up their distances from other members using 3 plots instead of 2.

Kurd
03-24-2015, 04:55 PM
So my closest match is my aunt. I would have expected some of the Armenians to match with Humanist. Like Armenian 6, Armenian 4 and Armenian 5 have higher southwest Asian and depending on the calculator match with Assyrians first.



I suspect that it has to to do with the Complete linkage (furthest neighbor) method used to analyze the clusters. In this method, the distances between clusters are determined by the greatest distance between any two objects in the different clusters (i.e., by the "furthest neighbors").

This may change when I redo the analysis later using one of the linkage methods I described above :)

Arbogan
03-24-2015, 06:00 PM
I must be either Bactrian or Sogtrian with the amount of west-asian in me ............or should I say.... tajik/uzbek? Its just farmer ancestry , in this case. Indo iranian usually is SWC with something steppe euro-like. Kurd i wonder if you could add uyghur to see their comparative scores with other indo-iranian dervatives.

Kurd
03-24-2015, 06:38 PM
Its just farmer ancestry , in this case. Indo iranian usually is SWC with something steppe euro-like. Kurd i wonder if you could add uyghur to see their comparative scores with other indo-iranian dervatives.

Ok, but I need their data.

Kurd
03-25-2015, 04:25 PM
The following are dendograms using the Eucladian distance method to determine distance from other members. In the 1st dendogram, I have used the Furthest Neighbor method for cluster linking (see previous post for explanation). Bol Nat, Pak Gujjar, Zeph, and 3 southern Kurds (KD...) have been added (courtesy Palisto). If you were not added, it probably is because you had posted or PMed your data in the wrong format (see posts in the beginning of thread). I have also posted the clusteroid info table for those interested.

In the 2nd dendogram, the Group Average method for cluster linking was used. This resulted in changes to a couple of cluster placements. For example, Jesus and my cluster was shifted to the west asian side from the S/SC Asian side. This is because, using the 1st method, which uses furthest neighbor, my distance to the most NW shifted individual in the Iranian cluster was greater than Jesus's distance to the most SE shifted member in the Iranian Baloch cluster (Baloch 3), thus we were placed in the S Asian cluster. Using the 2nd method, the average of my and Jesus's cluster was closer to the average of the Iranian cluster than the average of the S Asian cluster, thus we were placed in the West Asian cluster.


http://i1377.photobucket.com/albums/ah42/dil88888/Community%20Dendogram%20Furthest%20Neighbor%20Eucl adian_zpsqr4jnqh1.jpg (http://s1377.photobucket.com/user/dil88888/media/Community%20Dendogram%20Furthest%20Neighbor%20Eucl adian_zpsqr4jnqh1.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/Community%20Dendogram%20Group%20Averages%20Eucladi an_zpsuz1e6m1o.jpg (http://s1377.photobucket.com/user/dil88888/media/Community%20Dendogram%20Group%20Averages%20Eucladi an_zpsuz1e6m1o.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/Community%20Dendogram%20Group%20Averages%20Eucladi an%20Clusteroid%20info_zpsczwblymf.jpg (http://s1377.photobucket.com/user/dil88888/media/Community%20Dendogram%20Group%20Averages%20Eucladi an%20Clusteroid%20info_zpsczwblymf.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/Community%20Dendogram%20Furthest%20Neighbor%20Eucl adian%20Steps_zpsryaswaze.jpg (http://s1377.photobucket.com/user/dil88888/media/Community%20Dendogram%20Furthest%20Neighbor%20Eucl adian%20Steps_zpsryaswaze.jpg.html)

surbakhunWeesste
03-25-2015, 05:10 PM
http://i1377.photobucket.com/albums/ah42/dil88888/Community%20Dendogram%20Furthest%20Neighbor%20Eucl adian_zpsqr4jnqh1.jpg (http://s1377.photobucket.com/user/dil88888/media/Community%20Dendogram%20Furthest%20Neighbor%20Eucl adian_zpsqr4jnqh1.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/Community%20Dendogram%20Group%20Averages%20Eucladi an_zpsuz1e6m1o.jpg (http://s1377.photobucket.com/user/dil88888/media/Community%20Dendogram%20Group%20Averages%20Eucladi an_zpsuz1e6m1o.jpg.html)


Interesting, I thought Rukha would me in my place and vice-versa,since he is Tajik-Pashtun, that too from a different tribe. I wish we had data of Pashtuns from Northern and North-Eastern Afghanistan. I think the Pashtuns from North-Eastern Afghanistan will cluster closer to AdamYZ and Sein due to historical ties.

Kurd
03-25-2015, 05:24 PM
Interesting, I thought Rukha would me in my place and vice-versa,since he is Tajik-Pashtun, that too from a different tribe. I wish we had data of Pashtuns from Northern and North-Eastern Afghanistan. I think the Pashtuns from North-Eastern Afghanistan will cluster closer to AdamYZ and Sein due to historical ties.

The program compares admixture component by component for all 22 components. I may have your data switched with Kandhari. I meant to ask you earlier, but forgot. Who is the one with higher S Asian, or higher European?

surbakhunWeesste
03-25-2015, 05:39 PM
The program compares admixture component by component for all 22 components. I may have your data switched with Kandhari. I meant to ask you earlier, but forgot. Who is the one with higher S Asian, or higher European?
I have higher S.Asian and European as well.

Rukha
03-25-2015, 08:05 PM
Interesting, I thought Rukha would me in my place and vice-versa,since he is Tajik-Pashtun, that too from a different tribe. I wish we had data of Pashtuns from Northern and North-Eastern Afghanistan. I think the Pashtuns from North-Eastern Afghanistan will cluster closer to AdamYZ and Sein due to historical ties.

Pashtuns from eastern provinces like Paktia probably cluster closest to them. I think my mom (Wardaki) would cluster closer to them as well based on my results.

DIY oracle:
[6,] "43.3% Pakistani_Pushtun + 56.7% Tajik_Pomiri" "3.5459"

Oracle 4 on gedmatch:
Using 2 populations approximation:
1 50% Pakistani_Pushtun +50% Tajik_Pomiri_Shugnan @ 4.334452

Sein
03-25-2015, 08:10 PM
Pashtuns from eastern provinces like Paktia probably cluster closest to them. I think my mom (Wardaki) would cluster closer to them as well based on my results.

DIY oracle:
[6,] "43.3% Pakistani_Pushtun + 56.7% Tajik_Pomiri" "3.5459"

Oracle 4 on gedmatch:
Using 2 populations approximation:
1 50% Pakistani_Pushtun +50% Tajik_Pomiri_Shugnan @ 4.334452

True, the HGDP Pashtuns should be very representative of Pashtuns from Paktia and Khost, as well as Karlani Pashtuns from northern Paktika. Myself and Adam are probably closest to Pashtuns from Nangarhar, Laghman, and Kunar.

surbakhunWeesste
03-25-2015, 08:22 PM
Pashtuns from eastern provinces like Paktia probably cluster closest to them. I think my mom (Wardaki) would cluster closer to them as well based on my results.

DIY oracle:
[6,] "43.3% Pakistani_Pushtun + 56.7% Tajik_Pomiri" "3.5459"

Oracle 4 on gedmatch:
Using 2 populations approximation:
1 50% Pakistani_Pushtun +50% Tajik_Pomiri_Shugnan @ 4.334452

Ha! I didn't notice your 50-50 oracle. Nice.

The Barnacle
03-25-2015, 10:23 PM
The program compares admixture component by component for all 22 components. I may have your data switched with Kandhari. I meant to ask you earlier, but forgot. Who is the one with higher S Asian, or higher European?

I'm so confused, why does viktor and hazara ckuster closer to west Asians than afghans?

The Barnacle
03-25-2015, 10:25 PM
True, the HGDP Pashtuns should be very representative of Pashtuns from Paktia and Khost, as well as Karlani Pashtuns from northern Paktika. Myself and Adam are probably closest to Pashtuns from Nangarhar, Laghman, and Kunar.

It would be really interesting too see resujts from laghman, since many Tajiks have been pashtunized over there. I'm more curious about Western Afghanistan, there's not one single western afghan results.

The Barnacle
03-25-2015, 10:27 PM
Pashtuns from eastern provinces like Paktia probably cluster closest to them. I think my mom (Wardaki) would cluster closer to them as well based on my results.

DIY oracle:
[6,] "43.3% Pakistani_Pushtun + 56.7% Tajik_Pomiri" "3.5459"

Oracle 4 on gedmatch:
Using 2 populations approximation:
1 50% Pakistani_Pushtun +50% Tajik_Pomiri_Shugnan @ 4.334452

We'll tbh that's assuming your panjshiri side is identical to a pamiri...I reckon panjshiris will be similar to Northern Pashtuns and Farsiwans. You should get both of your parents tested.

surbakhunWeesste
03-26-2015, 01:52 AM
We'll tbh that's assuming your panjshiri side is identical to a pamiri...I reckon panjshiris will be similar to Northern Pashtuns and Farsiwans. You should get both of your parents tested.

How about you? Have you been tested? It would be interesting to see your results.

Kurd
03-26-2015, 03:14 AM
I'm so confused, why does viktor and hazara ckuster closer to west Asians than afghans?

I believe you are referring to the Group Averages plot. I will give an overview of how to read my plot, which should help with your question. This should also help others better understand it:

http://i1377.photobucket.com/albums/ah42/dil88888/Community%20Group%20Average%20Dendogram%20lesson_z psvl8wvcf4.jpg (http://s1377.photobucket.com/user/dil88888/media/Community%20Group%20Average%20Dendogram%20lesson_z psvl8wvcf4.jpg.html)

So basically the plot is showing that the distance from Viktor & Hazara's cluster to the W Asian cluster is 0.38 (arrow 4), and the distance to the S/SC Asian cluster is 0.43. What this means is that the average of Viktor & Hazara's admixture components is more overlapping with the average of the W Asian cluster (average of Armenian, Turk, Assyrian, and Iranian sub-clusters).

Similarly, the distance of the SC sub-cluster containing my daughter, Rukha, Kandhari, and Zahra to the Iranian Baloch cluster is 0.13 (arrow 3), whereas it is 0.26 from the average of the S. Asian clusters (arrow 2).

The distance between the average of the north S Asian cluster to the south S Asian cluster is 0.18 (arrow 1). The distance from the average of my and Jesus's subcluster to the average of the Iranian/Kurd cluster is 0.125 (arrow 6).

I will be explaining the other dendogram (Furthest Neighbor cluster linkage) in a subsequent post).

I will also be working on updated PCAs later, including an Asian PCA, which should include data on a group of W Asian individuals that I recently received.




EDIT: I am also seeking MDLP data on Parsis, Uyghurs, and western Afghans to include in this project.

jesus
03-26-2015, 04:15 AM
Pashtuns from eastern provinces like Paktia probably cluster closest to them. I think my mom (Wardaki) would cluster closer to them as well based on my results.

DIY oracle:
[6,] "43.3% Pakistani_Pushtun + 56.7% Tajik_Pomiri" "3.5459"

Oracle 4 on gedmatch:
Using 2 populations approximation:
1 50% Pakistani_Pushtun +50% Tajik_Pomiri_Shugnan @ 4.334452

Is the Pakisatni pashtun sample in MDLP k23b similar to the HGDP Pathans (also Pakistani Pashtun) ?
You seem to be closer to Pakistani Pashtuns than Pathans, I wonder where were those Pakistani Pashtuns from.
http://www.anthrogenica.com/showthread.php?4009-MDLP-K23b-Project-Part-2-of-2-Community-PCAs-graphs-amp-charts&p=74013&viewfull=1#post74013

Rukha
03-26-2015, 04:42 AM
We'll tbh that's assuming your panjshiri side is identical to a pamiri...I reckon panjshiris will be similar to Northern Pashtuns and Farsiwans. You should get both of your parents tested.

Well I'm just speculating on the basis of geographic proximity. Panjsher and Badakhshan (where Pamiris are concentrated) border each other. Tajiks from those provinces along with Parwan, Baghlan, Takhar, Kapisa, Kohistan (north of Kabul), etc. are quite similar culturally and historically...I'd expect Panjshiris to be more similar to them genetically than to Farsiwans. Who do you mean by northern Pashtuns?


Is the Pakisatni pashtun sample in MDLP k23b similar to the HGDP Pathans (also Pakistani Pashtun) ?
You seem to be closer to Pakistani Pashtuns than Pathans, I wonder where were those Pakistani Pashtuns from.
http://www.anthrogenica.com/showthre...ll=1#post74013

Not sure..Sein might know. I guess it's a less South Indian shifted sample than the HGDP Pashtuns.

The Barnacle
03-26-2015, 10:34 AM
I believe you are referring to the Group Averages plot. I will give an overview of how to read my plot, which should help with your question. This should also help others better understand it:

http://i1377.photobucket.com/albums/ah42/dil88888/Community%20Group%20Average%20Dendogram%20lesson_z psvl8wvcf4.jpg (http://s1377.photobucket.com/user/dil88888/media/Community%20Group%20Average%20Dendogram%20lesson_z psvl8wvcf4.jpg.html)

So basically the plot is showing that the distance from Viktor & Hazara's cluster to the W Asian cluster is 0.38 (arrow 4), and the distance to the S/SC Asian cluster is 0.43. What this means is that the average of Viktor & Hazara's admixture components is more overlapping with the average of the W Asian cluster (average of Armenian, Turk, Assyrian, and Iranian sub-clusters).

Similarly, the distance of the SC sub-cluster containing my daughter, Rukha, Kandhari, and Zahra to the Iranian Baloch cluster is 0.13 (arrow 3), whereas it is 0.26 from the average of the S. Asian clusters (arrow 2).

The distance between the average of the north S Asian cluster to the south S Asian cluster is 0.18 (arrow 1). The distance from the average of my and Jesus's subcluster to the average of the Iranian/Kurd cluster is 0.125 (arrow 6).

I will be explaining the other dendogram (Furthest Neighbor cluster linkage) in a subsequent post).

I will also be working on updated PCAs later, including an Asian PCA, which should include data on a group of W Asian individuals that I recently received.




EDIT: I am also seeking MDLP data on Parsis, Uyghurs, and western Afghans to include in this project.

Okay that's much more understandable lol thanks Kurd :). What's the distancs between the south Central Asian ckuster to west Asian cluster?

Seeking as in you haven't found anyone yet? What's up with these heratis

The Barnacle
03-26-2015, 10:38 AM
Well I'm just speculating on the basis of geographic proximity. Panjsher and Badakhshan (where Pamiris are concentrated) border each other. Tajiks from those provinces along with Parwan, Baghlan, Takhar, Kapisa, Kohistan (north of Kabul), etc. are quite similar culturally and historically...I'd expect Panjshiris to be more similar to them genetically than to Farsiwans. Who do you mean by northern Pashtuns?



Not sure..Sein might know. I guess it's a less South Indian shifted sample than the HGDP Pashtuns.

Possibly... Northern Pashtuns as in kunduz, baghlan and laghman. My point is, that I don't think your mother who is a pashtun from wardak will be that similar to HGDP Pathans IMO, she's going to be more Similar to metspalu Pashtuns. Aren't you curious about Pashtuns and Tajiks from Farah, Herat and other western provinces? I'm so curious lol

The Barnacle
03-26-2015, 10:45 AM
How about you? Have you been tested? It would be interesting to see your results.

Hopefully I will, I'm still young( 18). I'm just do confused about our people sometimes, I know my paternal grandmother was a popalzai from Kandahar, and maternal grandmother is a Tajik from Kabul, but paternal grandfather was from laghman who spoke Farsi and Pashto yet called himself Tajik, but also said his forefathers came from paktia? He didn't really care about these stuff. My maternal grandfather is half pashtun( fermuli tribe from suburbs of Kabul?) but his father? I don't even know lol. If I do a test how would I contribute? I'm too mixed between Tajiks and Pashtuns. Your In a better positio since your fully kandahari.

Maybe I'm a good representation of afghans in the west lol since many are mixed

Kurd
03-26-2015, 12:55 PM
I revised the community PCA based on a new set of consolidated components, which caused the eigenvectors to line up nicely. SW Asia consists of Near East & N Africa. W Asian is Caucasus. NE Asia consists of all the Siberian & NE Asian comps. SE Asia includes Melanesian, Austronesian, Australoid. African contains all the African comps - N Africa. Anc. Altaic is stand alone. European is EEF + EHG.

I have also posted the Eigenvectors to give an idea of the component axes. The coefficient table shows that most components are well represented by PC1 and PC2.

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20COMMUNITY_zpskujfm5pt.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20COMMUNITY_zpskujfm5pt.jpg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20COMMUNITY%20EIGENVECTORS_zpsmvaule67.j pg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20COMMUNITY%20EIGENVECTORS_zpsmvaule67.j pg.html)

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20COMMUNITY%20EIGENVALUES%20amp%20PC%20C OEFF_zpsp4opxhpj.jpg (http://s1377.photobucket.com/user/dil88888/media/PC1%20PC2%20COMMUNITY%20EIGENVALUES%20amp%20PC%20C OEFF_zpsp4opxhpj.jpg.html)

Kurd
03-26-2015, 02:40 PM
Okay that's much more understandable lol thanks Kurd :). What's the distancs between the south Central Asian ckuster to west Asian

Undeterminable from either dendogram

vettor
03-26-2015, 05:08 PM
Undeterminable from either dendogram

thanks for all your data, I find it very interesting.

A question......do you see the uzbeks as a central-asian or west-asian group ?

I understand via history that Kazaks, Nogai, Sirir, Tajiks and others branched out from Uzbeks and are noted as central-asian ( in the majority) , but I feel Tajiks and "pure" uzbeks are west-asian. ..............does your testing show this?

DMXX
03-26-2015, 05:19 PM
What genetic data could you possibly base that opinion on?

Although there's evidently genetic overlap between South-Central Asia and West Asia, every single PCA plot featuring major Eurasian regional groups indicates Tajiks and Uzbeks belong to their own distinct biogeographic cluster separate from West Asia.

surbakhunWeesste
03-26-2015, 05:46 PM
Hopefully I will, I'm still young( 18). I'm just do confused about our people sometimes, I know my paternal grandmother was a popalzai from Kandahar, and maternal grandmother is a Tajik from Kabul, but paternal grandfather was from laghman who spoke Farsi and Pashto yet called himself Tajik, but also said his forefathers came from paktia? He didn't really care about these stuff. My maternal grandfather is half pashtun( fermuli tribe from suburbs of Kabul?) but his father? I don't even know lol. If I do a test how would I contribute? I'm too mixed between Tajiks and Pashtuns. Your In a better positio since your fully kandahari.

Maybe I'm a good representation of afghans in the west lol since many are mixed

Now, I am much interested to see your results, quite a blend :) . Also, there is no better or less position. I have few family members who are mixed with other Afghans and other nationalities as well and boy their results are fascinating.

surbakhunWeesste
03-26-2015, 06:05 PM
[QUOTE=vettor;76092]

A question......do you see the uzbeks as a central-asian or west-asian group ?

[/QUOTE ]

There are more Tajiks in Afghanistan than in Tajikistan. There are many Tajikified Uzbeks in Afghanistan as well. Are you taking about Tajiks and Uzbeks from Tajikistan and Uzbekistan respectively? Also, Tajik basily means Persian, so in that way of course all Tajiks are Persians and hence West Asians. There is a cultural backup to a certain degree as well. However, genetically, Tajiks from Tajikistan vs Tajiks from Afghanistan(based on which part) will show varying results (I have seen quite few results of friends and family).
Uzebks, are perhaps more central Asian than West-Asian, culturally and genetically. We need more genetic results ofc.

vettor
03-26-2015, 06:10 PM
What genetic data could you possibly base that opinion on?

Although there's evidently genetic overlap between South-Central Asia and West Asia, every single PCA plot featuring major Eurasian regional groups indicates Tajiks and Uzbeks belong to their own distinct biogeographic cluster separate from West Asia.

i did not mention genetics, I was speaking about the zones which incorporate west-asian or not

jesus
03-26-2015, 06:16 PM
i did not mention geneics, I was speaking about the zones which incorporate west-asian or not:amen:

Tajikistan and Uzbekistan belong to the southern part of Central Asia, The most "west asian" influenced culturally, historically and sometimes genetically part of Central asia.

The Barnacle
03-26-2015, 07:03 PM
What genetic data could you possibly base that opinion on?

Although there's evidently genetic overlap between South-Central Asia and West Asia, every single PCA plot featuring major Eurasian regional groups indicates Tajiks and Uzbeks belong to their own distinct biogeographic cluster separate from West Asia.

Certainly, although we don't really know where true afghan Tajiks( panjsher, Kabul, kapisa, parwan, badakhshan, Herat, Farah etc) will pan out. Would they be similar to tajikstani Tajiks? What about heratis? We need more sanples for sure before we could make assumptions however.

The Barnacle
03-26-2015, 07:07 PM
[QUOTE=vettor;76092]

A question......do you see the uzbeks as a central-asian or west-asian group ?

[/QUOTE ]

There are more Tajiks in Afghanistan than in Tajikistan. There are many Tajikified Uzbeks in Afghanistan as well. Are you taking about Tajiks and Uzbeks from Tajikistan and Uzbekistan respectively? Also, Tajik basily means Persian, so in that way of course all Tajiks are Persians and hence West Asians. There is a cultural backup to a certain degree as well. However, genetically, Tajiks from Tajikistan vs Tajiks from Afghanistan(based on which part) will show varying results (I have seen quite few results of friends and family).
Uzebks, are perhaps more central Asian than West-Asian, culturally and genetically. We need more genetic results ofc.

Have you seen afghan Tajiks results? Could you maybe post their results or describe it? There is definety persianzed Uzbeks In Afghanistan, I wonder why? There a few afghan Tajiks who strikingly Have members who can pass off as a hazara or Uzbek.the 4 metspalu afghan Tajiks resujts just don't make sense to me...how can afghan Tajiks be more eastern Eurasian than tajikstani tsjikd? How is they possibly other than the fact that maybe they had hazara or Uzbek ancestry. In that case, they weren't true afghan Tajiks.

The Barnacle
03-26-2015, 07:08 PM
Tajikistan and Uzbekistan belong to the southern part of Central Asia, The most "west asian" influenced culturally, historically and sometimes genetically part of Central asia.

I'm pretty sure it's afghanistan in terms of geography, culture and genetics lol.

The Barnacle
03-26-2015, 07:10 PM
Now, I am much interested to see your results, quite a blend :) . Also, there is no better or less position. I have few family members who are mixed with other Afghans and other nationalities as well and boy their results are fascinating.

Mixed with Tajiks? From where? What do their admixture proportions turn out to look like?

Kurd
03-26-2015, 07:15 PM
Here are the results of a Mumbai Parsi, who I will add to the PCA along with a SW Iranian individual later tonight. It is interesting that in spite of centuries of isolation from Iran, it appears that the Parsis retain considerable admixture overlap with Iranians, and that Iranians & Kurds show up high on their list of closest proxies (more Parsi results are needed for a more definitive conclusion). They show up as my number 2 proxy, and I believe number 3 or 4 proxy for Jesus using this calculator.



Admix Results (sorted):

# Population Percent
1 South_Central_Asian 29.77
2 Caucasian 29.50
3 South_Indian 16.27
4 Near_East 10.48
5 Ancestral_Altaic 3.64
6 European_Hunters_Gatherers 3.18
7 European_Early_Farmers 2.89
8 North_African 2.19


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Parsi @ 4.660089
2 Iranian @ 11.152367
3 Afghan_Pushtun @ 14.547330
4 Kurd_South @ 15.551110
5 Kurd_East @ 15.611875
6 Pakistani_Pushtun @ 16.223436
7 Pashtun_Afghani @ 17.764494
8 Kurd @ 18.300711
9 Tajik_Afghan @ 18.417702
10 Kurd_North @ 18.508699
11 Uzbek_Afghan @ 18.645628
12 Azeri @ 19.039072
13 Baku_WGA @ 19.185410
14 Uzbek_Tashkent @ 20.802530
15 Tajik_Pomiri_Ishkashim @ 21.347321
16 Tajik_Yagnobi @ 21.360611
17 Tajik_Pomiri_Rushan @ 22.043495
18 Ain_Touta_WGA @ 22.634521
19 Tajik_Pomiri_Shugnan @ 22.683516
20 Iraki @ 23.000721

Using 2 populations approximation:
1 50% Jew_Tat +50% Punjabi_Gujjar @ 2.560950


Using 3 populations approximation:
1 50% Iranian_Jew +25% Jatt_Pahari +25% Pathan @ 2.044134


Using 4 populations approximation:
1 GujaratiB_GIH + Kurd + Parsi + Uzbekistani_Jew @ 1.592891
2 Baku_WGA + GujaratiB_GIH + Parsi + Uzbekistani_Jew @ 1.678655
3 Baku_WGA + GujaratiA_GIH + Iraqi_Mandean + Parsi @ 1.706142
4 Iranian_Jew + Jew_Tat + Pathan + Punjabi_Gujjar @ 1.774127
5 Baku_WGA + Iraqi_Mandean + Pakistani + Parsi @ 1.814839
6 Assyrian_Arzni + Iranian_Jew + Pakistani + Pathan @ 1.815671
7 Jew_Tat + Parsi + Parsi + Pathan @ 1.841990
8 Assyrian_Arzni + Kurd_Jew + Pakistani + Pathan @ 1.845183
9 Kurd_East + Pakistani + Parsi + Uzbekistani_Jew @ 1.856343
10 Baku_WGA + GujaratiA_GIH + Kurd_Jew + Parsi @ 1.862647
11 Baku_WGA + Pakistani + Parsi + Uzbekistani_Jew @ 1.908549
12 Baku_WGA + GujaratiB_GIH + Kurd_Jew + Parsi @ 1.910076
13 Armenian_Yerevan + GujaratiA_GIH + Iranian + Parsi @ 1.910627
14 Kurd_South + Pakistani + Parsi + Uzbekistani_Jew @ 1.912251
15 Assyrian_Arzni + GujaratiA_GIH + Iranian + Parsi @ 1.923552
16 Pakistani_Pushtun + Parsi + Parsi + Uzbekistani_Jew @ 1.932260
17 Assyrian_Arzni + Iranian_Jew + Pathan + Punjabi_Gujjar @ 1.952825
18 Baku_WGA + GujaratiA_GIH + Iranian_Jew + Parsi @ 1.961465
19 Jew_Tat + Kurd_Jew + Pathan + Punjabi_Gujjar @ 1.961544
20 Armenian_Yerevan + Iranian_Jew + Pakistani + Pathan @ 1.976038

vettor
03-26-2015, 08:25 PM
[QUOTE=zahra;76101]

Have you seen afghan Tajiks results? Could you maybe post their results or describe it? There is definety persianzed Uzbeks In Afghanistan, I wonder why? There a few afghan Tajiks who strikingly Have members who can pass off as a hazara or Uzbek.the 4 metspalu afghan Tajiks resujts just don't make sense to me...how can afghan Tajiks be more eastern Eurasian than tajikstani tsjikd? How is they possibly other than the fact that maybe they had hazara or Uzbek ancestry. In that case, they weren't true afghan Tajiks.

No, I have not seen Tajiks results, have you got any, I am curious.

I am awaiting Havard analyis around the neighbourhood ( skeletral finds ) of Turkmenistan ( must be neolithic )

from havard below

The Venetians came from Turkmenistan?

The discovery of a bone plate in the oasis of Adji Kui confirms the hypothesis of archaeologists

The legendary Paphlagonia, mentioned by Homer and Strabo, really exists. The Venetian in all probability came from the legendary land, located not far from Turkmenistan in the Caspian Sea. The latest confirmation comes from the discovery of a bone plate of the third millennium BC in the course of an excavation campaign in the oasis of Adji Kui, not far from the Caspian Sea. The small object is decorated with carved rosettes, a pattern that is found frequently in the objects of culture paleoveneta. The legend of the origin of the Asian Veneto comes from Homer, quoted by Strabo, who speaks of Paphlagonia, homeland of the ancient people of the Eneti, or the Venetians. It 'a quarter of a century that the mission of Gabriele Rossi Osmida is bringing to light some towns belonging to the "oasis of civilization", the caravan in the desert. And slowly recovers the history of the suggestions of the literature and confirms the reliability of ancient traditions. The next step, in collaboration with the Department of Anthropology at the University of Harvard, is the map of the DNA of the skeletons found.

fil
03-26-2015, 09:05 PM
Quick question, why am I far from the other South Asians in the PCA?

jesus
03-26-2015, 09:38 PM
I'm pretty sure it's afghanistan in terms of geography, culture and genetics lol.

Of course Afghanistan has the most west asian influence in the southern parts of central Asia, but vettor was talking about Uzbekistan and Tajikistan. That's why I didn't mention it.

The Barnacle
03-26-2015, 09:57 PM
Quick question, why am I far from the other South Asians in the PCA?

I think the milkman is involved people

I'm joking lol liven up the thread man :)

surbakhunWeesste
03-26-2015, 10:13 PM
[QUOTE=The Barnacle;76112]

No, I have not seen Tajiks results, have you got any, I am curious.
[/B]


Mixed with Tajiks? From where? What do their admixture proportions turn out to look like?

I have seen Afghan Tajik admixture results, but I do not have them with me sorry.
Pashtun father + Tajik mother both are Kandaharis, however the mother's ancestral family originally hails from Herat. I don't remember the detailed admixture proportion but I will ask them for it. They are not much curious about genetics.

The Barnacle
03-26-2015, 10:18 PM
[QUOTE=vettor;76127]



I have seen Afghan Tajik admixture results, but I do not have them with me sorry.
Pashtun father + Tajik mother both are Kandaharis, however the mother's ancestral family originally hails from Herat. I don't remember the detailed admixture proportion but I will ask them for it. They are not much curious about genetics.

If they ain't curious why did they take a test Lool? Herati? That's a first, please do Zahra thanks

surbakhunWeesste
03-26-2015, 10:20 PM
[QUOTE=zahra;76138]

If they ain't curious why did they take a test Lool? Herati? That's a first, please do Zahra thanks

I made them, lolz. They only wanna identify as Afghans. Its worthless really for even stating the results unless I can provide them actually. SMH

The Barnacle
03-26-2015, 10:41 PM
[QUOTE=The Barnacle;76140]

I made them, lolz. They only wanna identify as Afghans. Its worthless really for even stating the results unless I can provide them actually. SMH

Attagirl loool.....please provide the results, I'm so intrigued to see a heratis results

Kurd
03-27-2015, 12:29 AM
Quick question, why am I far from the other South Asians in the PCA?
I dont have your numbers memorized, but look at the Eigenvector chart on the previous page. If you have more S Indian you will get pulled in a NE direction. If they have more Anc Altaic or NE Asian they will get pulled in a SE direction. Also look at the European vector. It also has some influence

fil
03-27-2015, 01:13 AM
I dont have your numbers memorized, but look at the Eigenvector chart on the previous page. If you have more S Indian you will get pulled in a NE direction. If they have more Anc Altaic or NE Asian they will get pulled in a SE direction. Also look at the European vector. It also has some influence

Thank you!! I basically lack those components. I guess that just means I am uber South Indian.

Kurd
03-27-2015, 04:01 AM
The PCA has been updated with the addition of a Mumbai Parsi, PARSI1, and an Iranian Shirazi, SHIRAZI1. I was surprised to see that PARSI1 plotted right next to me, indicating a high admixture overlap (surprising considering the isolation of Parsis from Iran for the past `1000years). SHIRAZI1 plotted right next to Jesus.



http://i.imgur.com/Ps245CP.jpg


Admix Results (sorted) for PARSI1:

# Population Percent
1 South_Central_Asian 29.77
2 Caucasian 29.50
3 South_Indian 16.27
4 Near_East 10.48
5 Ancestral_Altaic 3.64
6 European_Hunters_Gatherers 3.18
7 European_Early_Farmers 2.89
8 North_African 2.19


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Parsi @ 4.660089
2 Iranian @ 11.152367
3 Afghan_Pushtun @ 14.547330
4 Kurd_South @ 15.551110
5 Kurd_East @ 15.611875
6 Pakistani_Pushtun @ 16.223436
7 Pashtun_Afghani @ 17.764494
8 Kurd @ 18.300711
9 Tajik_Afghan @ 18.417702
10 Kurd_North @ 18.508699
11 Uzbek_Afghan @ 18.645628
12 Azeri @ 19.039072
13 Baku_WGA @ 19.185410
14 Uzbek_Tashkent @ 20.802530
15 Tajik_Pomiri_Ishkashim @ 21.347321
16 Tajik_Yagnobi @ 21.360611
17 Tajik_Pomiri_Rushan @ 22.043495
18 Ain_Touta_WGA @ 22.634521
19 Tajik_Pomiri_Shugnan @ 22.683516
20 Iraki @ 23.000721

Using 2 populations approximation:
1 50% Jew_Tat +50% Punjabi_Gujjar @ 2.560950


Using 3 populations approximation:
1 50% Iranian_Jew +25% Jatt_Pahari +25% Pathan @ 2.044134


Using 4 populations approximation:
1 GujaratiB_GIH + Kurd + Parsi + Uzbekistani_Jew @ 1.592891
2 Baku_WGA + GujaratiB_GIH + Parsi + Uzbekistani_Jew @ 1.678655
3 Baku_WGA + GujaratiA_GIH + Iraqi_Mandean + Parsi @ 1.706142
4 Iranian_Jew + Jew_Tat + Pathan + Punjabi_Gujjar @ 1.774127
5 Baku_WGA + Iraqi_Mandean + Pakistani + Parsi @ 1.814839
6 Assyrian_Arzni + Iranian_Jew + Pakistani + Pathan @ 1.815671
7 Jew_Tat + Parsi + Parsi + Pathan @ 1.841990
8 Assyrian_Arzni + Kurd_Jew + Pakistani + Pathan @ 1.845183
9 Kurd_East + Pakistani + Parsi + Uzbekistani_Jew @ 1.856343
10 Baku_WGA + GujaratiA_GIH + Kurd_Jew + Parsi @ 1.862647

Kurd
03-27-2015, 05:12 AM
The dendogram has been updated with the addition of a Mumbai Parsi, and a Shirazi from Iran. Both use the Euclidean method for member to member distance measurement. The 1st one uses the "Furthest Neighbor" cluster linkage method. The second uses "Group Average"


http://i.imgur.com/eecr2XF.jpg

http://i.imgur.com/V6iV3nA.jpg

Hanna
03-27-2015, 07:18 AM
The dendogram has been updated with the addition of a Mumbai Parsi, and a Shirazi from Iran. Both use the Euclidean method for member to member distance measurement. The 1st one uses the "Furthest Neighbor" cluster linkage method. The second uses "Group Average"


http://i.imgur.com/eecr2XF.jpg

http://i.imgur.com/V6iV3nA.jpg
Why is it that in the dendograms I branch out with Armenians but in the PCAs I plot with the other Turkish members?

paulgill
03-27-2015, 07:27 AM
Here is mine K23b results.

Population
Amerindian -
Ancestral_Altaic 6.67%
South_Central_Asian 37.06%
Arctic 2.11%
South_Indian 31.40%
Australoid 0.28%
Austronesian 0.38%
Caucasian 11.64%
Archaic_Human -
East_African -
East_Siberian 0.10%
European_Early_Farmers 0.88%
Khoisan -
Melano_Polynesian 0.24%
Archaic_African -
Near_East 2.37%
North_African -
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 6.86%

Kurd
03-27-2015, 12:56 PM
The PCA has been updated with the addition of a Mumbai Parsi, PARSI1, and an Iranian Shirazi, SHIRAZI1. I was surprised to see that PARSI1 plotted right next to me, indicating a high admixture overlap (surprising considering the isolation of Parsis from Iran for the past `1000years). SHIRAZI1 plotted right next to Jesus.



http://i.imgur.com/Ps245CP.jpg


Admix Results (sorted) for PARSI1:

# Population Percent
1 South_Central_Asian 29.77
2 Caucasian 29.50
3 South_Indian 16.27
4 Near_East 10.48
5 Ancestral_Altaic 3.64
6 European_Hunters_Gatherers 3.18
7 European_Early_Farmers 2.89
8 North_African 2.19


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Parsi @ 4.660089
2 Iranian @ 11.152367
3 Afghan_Pushtun @ 14.547330
4 Kurd_South @ 15.551110
5 Kurd_East @ 15.611875
6 Pakistani_Pushtun @ 16.223436
7 Pashtun_Afghani @ 17.764494
8 Kurd @ 18.300711
9 Tajik_Afghan @ 18.417702
10 Kurd_North @ 18.508699
11 Uzbek_Afghan @ 18.645628
12 Azeri @ 19.039072
13 Baku_WGA @ 19.185410
14 Uzbek_Tashkent @ 20.802530
15 Tajik_Pomiri_Ishkashim @ 21.347321
16 Tajik_Yagnobi @ 21.360611
17 Tajik_Pomiri_Rushan @ 22.043495
18 Ain_Touta_WGA @ 22.634521
19 Tajik_Pomiri_Shugnan @ 22.683516
20 Iraki @ 23.000721



The Community PCA PC3 PC4 has been updated with the addition of Parsi & Shirazi. I apologize that I had to turn off quite a few name labels to make it legible. I had no choice if I wanted to keep Ust Ishim on the PCA. I will be doing a zoomed in version without Ust Ishim later, and will be able to turn on some name labels. You will note from the eigenvectors that SE/NE Asian and African components are better represented here than in PC1 PC2 PCA.

The importance of looking at both PCAs simultaneously to judge distance from other members is shown here. In PC1 PC2 I plotted almost on top of Parsi, whereas here I am a little distant from him (due to my slightly higher SE Asian & African comps., see eigenvectors).Parsi still appears to have the highest admixture overlap with me, followed by Jesus, according to the dendogram. To gauge distance, I would have to add my distance from PC1 PC2 PCA to the PC3 PC4 PCA. The same would apply to all other members.


http://i.imgur.com/MP2EghM.jpg

http://i.imgur.com/doGZQsO.jpg

http://i.imgur.com/y6i9j6V.jpg

Kurd
03-27-2015, 01:49 PM
Why is it that in the dendograms I branch out with Armenians but in the PCAs I plot with the other Turkish members?

The reason is that I had to remove your closest proxies, Turk HA, and another Turkish member from the cluster analysis spreadsheet for legibility reasons. That left you with Armenian 6 as your closest admixture overlapping proxy :)

Hanna
03-27-2015, 03:16 PM
The reason is that I had to remove your closest proxies, Turk HA, and another Turkish member from the cluster analysis spreadsheet for legibility reasons. That left you with Armenian 6 as your closest admixture overlapping proxy :)

But even in the previous dendograms and PCAs you made where my aunt ( Turk HA ) and the other Turkish members were included the situation was the same. I plotted with Turkish members and branched with the Armenians.

Like here:

http://i1377.photobucket.com/albums/ah42/dil88888/Community%20Group%20Average%20Dendogram%20lesson_z psvl8wvcf4.jpg

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20COMMUNITY_zpskujfm5pt.jpg

Kurd
03-27-2015, 04:39 PM
But even in the previous dendograms and PCAs you made where my aunt ( Turk HA ) and the other Turkish members were included the situation was the same. I plotted with Turkish members and branched with the Armenians.

Like here:

http://i1377.photobucket.com/albums/ah42/dil88888/Community%20Group%20Average%20Dendogram%20lesson_z psvl8wvcf4.jpg

http://i1377.photobucket.com/albums/ah42/dil88888/PC1%20PC2%20COMMUNITY_zpskujfm5pt.jpg

I think your question is why are you plotting with Turks on PC1/PC2 and branching with Armenians. There are a few things affecting the outcome in your case:

1- In the first go around, your aunt was included in the cluster spreadsheet. The consequences of your aunt being there on the spreadsheet caused the following:
a- Being that your aunt is Armenian shifted, it caused the group average of your cluster (you & your aunt) to get Armenian shifted. When
b- The program did its comparisons it found the avg of your cluster was closer to Armenian6, than to the average of the turkish cluster. In effect your aunt dragged you to Armenian6. If she had not been on the spreadsheet the outcome may have been different.

2- In the 2nd round, the only Turks in the cluster analysis spreadsheet were Miss..., Afshar, and Anabasis. If you look at both PCAs you will note that Miss.. & Anabasis are quite distant from you on PC1/PC2. Afshar is close to you. However when you look at PC3/PC4, he is not too close (you probably cant see him since his name label is missing). Remember to add distances from both PCAs. So when the program does the comparison, it decides to pair Afshar with Miss..., since he is a little closer to him, and you are left with Armenian6.

Hanna
03-27-2015, 05:05 PM
I think your question is why are you plotting with Turks on PC1/PC2 and branching with Armenians. There are a few things affecting the outcome in your case:

Very well explained and understood. My mother being Armenian shifted and father being average Turk makes me in between the two all the time.Thanks for your observation.

paulgill
03-27-2015, 10:07 PM
What I can't understand is you adding South Central Asian to South Indian. South Central Asian is close to Gedrosia which is basicaly Caucasus+a little bit of West Asia.

Darko
03-27-2015, 10:53 PM
Here are my results;
Amerindian -
Ancestral_Altaic -
South_Central_Asian 0.61%
Arctic -
South_Indian 1.45%
Australoid 0.23%
Austronesian -
Caucasian 6.16%
Archaic_Human -
East_African 11.84%
East_Siberian -
European_Early_Farmers 6.92%
Khoisan -
Melano_Polynesian -
Archaic_African 11.34%
Near_East 14.48%
North_African 12.58%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian 34.36%
Tungus-Altaic -
European_Hunters_Gatherers -

Kurd
03-28-2015, 12:44 AM
What I can't understand is you adding South Central Asian to South Indian. South Central Asian is close to Gedrosia which is basicaly Caucasus+a little bit of West Asia.

Where did you see that I added the two?

Varun R
03-28-2015, 01:10 AM
Kurd,

To answer your question from a while back, most of my ancestry would be from Karnataka. Granted, my mom's parents are both Iyers, with one parent ancestrally hailing from Kerala (Palakkad) and the other parent hailing from central Tamil Nadu (Cauvery River region). I recently got a new match on 23andMe. He is a Karnataka Brahmin from Udupi, and his K23 results are as follows:

Amerindian 1.68%
Ancestral_Altaic 1.93%
South_Central_Asian 32.40%
Arctic 0.73%
South_Indian 47.38%
Australoid -
Austronesian 0.99%
Caucasian 9.78%
Archaic_Human 0.13%
East_African -
East_Siberian 0.58%
European_Early_Farmers -
Khoisan -
Melano_Polynesian -
Archaic_African -
Near_East 0.89%
North_African 0.25%
Paleo_Siberian 0.18%
African_Pygmy -
South_East_Asian -
Subsaharian 0.80%
Tungus-Altaic -
European_Hunters_Gatherers 2.27%

The strange thing is, when I check his results using other admixture calculators, (ie Harappa) he obtains unusually high SW Asian (rather like Fil). He does not show MidEast on 23andMe's Ancestry Composition. His Harappa results:

# Population Percent
1 S-Indian 44.71
2 Baloch 36.43
3 Caucasian 7.69
4 NE-Euro 3.74
5 SW-Asian 3.49
6 Beringian 1.47
7 SE-Asian 0.98
8 American 0.53
9 Mediterranean 0.5
10 Siberian 0.31
11 NE-Asian 0.14
12 Papuan 0.01

Mixed mode oracle:
# Primary Population (source) Secondary Population (source) Distance
1 95.2% tn-brahmin (xing) + 4.8% yemen-jew (behar) @ 1.87
2 95.2% brahmin-tamil-nadu (metspalu) + 4.8% yemen-jew (behar) @ 1.93
3 95.3% brahmin-tamil-nadu (metspalu) + 4.7% saudi (behar) @ 1.96
4 95.3% tn-brahmin (xing) + 4.7% saudi (behar) @ 1.99
5 61.9% sindhi (hgdp) + 38.1% hakkipikki (metspalu) @ 2.03
6 55.9% tamil-nadu-scheduled-caste (metspalu) + 44.1% sindhi (harappa) @ 2.08
7 95.3% brahmin-tamil-nadu (metspalu) + 4.7% bedouin (hgdp) @ 2.09
8 95.1% tn-brahmin (xing) + 4.9% samaritian (behar) @ 2.13
9 94.9% brahmin-tamil-nadu (metspalu) + 5.1% qatari (henn2012) @ 2.13
10 94.7% tn-brahmin (xing) + 5.3% palestinian (hgdp) @ 2.15

I get similar results from Dodecad V3
# Primary Population (source) Secondary Population (source) Distance
1 94.5% TN_Brahmin (Xing) + 5.5% Iranians (Behar) @ 1.39
2 94.5% TN_Brahmin (Xing) + 5.5% Kurd (Dodecad) @ 1.44
3 94.8% TN_Brahmin (Xing) + 5.2% Uzbekistan_Jews (Behar) @ 1.46
4 95.2% TN_Brahmin (Xing) + 4.8% Georgia_Jews (Behar) @ 1.47
5 86.7% AP_Brahmin (Xing) + 13.3% Brahui (HGDP) @ 1.51
6 95.2% TN_Brahmin (Xing) + 4.8% Syrians (Behar) @ 1.51
7 95.2% TN_Brahmin (Xing) + 4.8% Azerbaijan_Jews (Behar) @ 1.52
8 94.8% TN_Brahmin (Xing) + 5.2% Iranian (Dodecad) @ 1.54
9 95.1% TN_Brahmin (Xing) + 4.9% Iranian_Jews (Behar) @ 1.54
10 95.3% TN_Brahmin (Xing) + 4.7% Iraq_Jews (Behar) @ 1.55

Kurd
03-28-2015, 01:37 AM
What I can't understand is you adding South Central Asian to South Indian. South Central Asian is close to Gedrosia which is basicaly Caucasus+a little bit of West Asia.

This statement is not very accurate either. Caucasus is included in the definition of W Asian, and is a distinct component from Near East (SW Asia), and SC Asia in the MDLP k23 calculator, or do you mean that the alleles associated with the SC Asian component are derived from alleles associated with the SW Asian or W Asian component?

Also, I checked my posts and did not find anywhere SC Asian added to S Indian. In fact, if you check the eigenvectors or eigenvalues associated with the PCAs you will see that they are not combined.

rms2
03-28-2015, 01:42 AM
I hope it is not too off-topic for me to stop momentarily and offer a hearty thanks to Kurd for including me in his updated plots, charts, and graphs, etc. Very cool! Thanks!

I had kind of given up on being included (my fault entirely - I was late getting involved in it) and had not checked this thread in a few days. I was surprised and pleased to see myself included in the community results.

paulgill
03-28-2015, 12:40 PM
"The following definitions will apply throughout the project:

1- S. Asia = SC Asia + S. India
2- NE & SE Asia = The sum of Amerindian, Ancestral Altaic, Arctic, Tungus Altaic, Paleo Siberian, SE Asian, Australoid, Austronesian, Melano Polynesian
3- W. Asia = Near East + Caucasian
4- The rest is self explanatory"

May be I have misunderstood you.

Kurd
03-28-2015, 09:40 PM
"The following definitions will apply throughout the project:

1- S. Asia = SC Asia + S. India
2- NE & SE Asia = The sum of Amerindian, Ancestral Altaic, Arctic, Tungus Altaic, Paleo Siberian, SE Asian, Australoid, Austronesian, Melano Polynesian
3- W. Asia = Near East + Caucasian
4- The rest is self explanatory"

May be I have misunderstood you.

That was my mistake, sorry. I did not combine them in my PCAs or other graphs. Although somwhat positively correlated with each other as per the eigenvectors, there is distiction between the two.

Kurd
03-28-2015, 11:33 PM
For those wondering, a heads up, I am working on a PCA tailored to the W & S Asian community. I will attempt to break it further down by ethnicities than the previous PCAs. The geographic coverage will be limited to W and S Asia to keep it legible because of the number of individuals. It should be uploaded by tomorrow sometime.

paulgill
03-29-2015, 01:26 AM
Kurd, who are these people who use Khabur as surname?

Kurd
03-29-2015, 02:42 AM
Kurd, who are these people who use Khabur as surname?

Personally I have never heard anyone use that name. I know it is a river in Syria. Sorry, my knowledge is quiet limited on matters west of Iran/Iraq

paulgill
03-29-2015, 05:27 AM
The little Khabur River is in Iraq and Turkey goes through Zakho. My Ydna lineage is J1-Z1853* which is supposedly from the Van lake area, my people use Khabra as their surname and their town name. I don't know when my ancestors left the area, it could be very long time ago, but they have held on to the name, it in my opinion have deep connection to Eber, Hebrew or Khabur River.

Kurd
03-29-2015, 08:42 AM
The little Khabur River is in Iraq and Turkey goes through Zakho. My Ydna lineage is J1-Z1853* which is supposedly from the Van lake area, my people use Khabra as their surname and their town name. I don't know when my ancestors left the area, it could be very long time ago, but they have held on to the name, it in my opinion have deep connection to Eber, Hebrew or Khabur River.

That is very interesting. I know that many Baloch tribes claim Kurd ancestry from various Kurdi tribes (Iran), and I believe the Rind Baloch (presentl living from Makran to Sibi in Pakistan) claim ancestry from Kaldani Kurds from around Halab, Syria.

paulgill
03-29-2015, 09:03 AM
That is very interesting. I know that many Baloch tribes claim Kurd ancestry from various Kurdi tribes (Iran), and I believe the Rind Baloch (presentl living from Makran to Sibi in Pakistan) claim ancestry from Kaldani Kurds from around Halab, Syria. Yes, I was aware of the Baloch claims especially to NW Iran, also to Syria, specifically from Kurd Tribes, this part I didn't know till now. Thanks.

Kurd
03-30-2015, 03:11 AM
These PCAs are dedicated to the Asian community at Anthrogenica. This series plots PC1 & PC2. You will note from the component coefficient table that most MDLP components are well represented by PC1 & PC2, except for SE/NE Asia and Africa. I will be uploading the plots for PC3 & PC4 later. Kurds are courtesy Palisto.

I strongly suggest you take a look at the eigenvectors to determine component axes on the PCA. This will explain the location of the members on the plot.

To clarify, since there was a little confusion, the PCA component assignments are as follows:

SW Asia = Near East + N Africa
W Asia = Caucasus
NE Asia= All Siberian comps
Anc. Altaic = Anc. Altaic
Africa = All African comps except N Africa
SE Asia = Includes Melanesian, Austronesian, Australoid
European = EEF + EHG

The 1st PCA is an overview, and subsequent ones zoom in on particular regions.



http://i.imgur.com/AUqG7dr.jpg

http://i.imgur.com/Xd6ZAOc.jpg

http://i.imgur.com/UmpltyE.jpg

http://i.imgur.com/hz5KKNJ.jpg

http://i.imgur.com/WP4A1R6.jpg

ViktorL1
03-30-2015, 03:23 AM
Kurd, thanks for the post, very informative PCA.

Is the community member indicated as "Tajik" a Tajikistani or Afghanistani Tajik?

Kurd
03-30-2015, 03:31 AM
Kurd, thanks for the post, very informative PCA.

Is the community member indicated as "Tajik" a Tajikistani or Afghanistani Tajik?

He was supplied to me by member Bol Nat. He is Afghan, and judging from his S Indian score of 15, and his best proxy of Tajik @8.5, I presumed he was an Afghan Tajik, although his distance of 8.5 may indicate a little non Tajik mixture, but that all depends on what Tajik references were used by the calculator

Edit: You may want to look at Rukha (Pashtun/Tajik) proxies and distances to get a better idea about Tajik's ancestry

Rukha
03-30-2015, 04:36 AM
He was supplied to me by member Bol Nat. He is Afghan, and judging from his S Indian score of 15, and his best proxy of Tajik @8.5, I presumed he was an Afghan Tajik, although his distance of 8.5 may indicate a little non Tajik mixture, but that all depends on what Tajik references were used by the calculator

Edit: You may want to look at Rukha (Pashtun/Tajik) proxies and distances to get a better idea about Tajik's ancestry

Interesting, would be nice to see his admixture scores and know his precise background.

parasar
03-30-2015, 04:46 AM
That is very interesting. I know that many Baloch tribes claim Kurd ancestry from various Kurdi tribes (Iran), and I believe the Rind Baloch (presentl living from Makran to Sibi in Pakistan) claim ancestry from Kaldani Kurds from around Halab, Syria.

See also the Dravidian speaking but otherwise similar Brahui:
https://books.google.com/books?id=_eqr833q9qYC&pg=PA32

Kurd
03-30-2015, 04:55 AM
Interesting, would be nice to see his admixture scores and know his precise background.

These were his scores. We are not sure what his ancestry is other than Afghan. I recall his closest proxy was Tajik @ 8.5. You may be in a better position to comment on his likely ancestry based on this info, and comparing to your scores :)

Amerindian 0.05%
Ancestral_Altaic 5.33%
South_Central_Asian 28.96%
Arctic 0
South_Indian 15.95%
Australoid 0.26%
Austronesian 0
Caucasian 23.44%
Archaic_Human 0
East_African 0.07%
East_Siberian 1.78%
European_Early_Farmers 3.71%
Khoisan 0
Melano_Polynesian 0
Archaic_African 0.17%
Near_East 4.24%
North_African 0
Paleo_Siberian 0
African_Pygmy 0
South_East_Asian 3.53%
Subsaharian 0
Tungus-Altaic 6.06%
European_Hunters_Gatherers 6.45%

Rukha
03-30-2015, 05:30 AM
These were his scores. We are not sure what his ancestry is other than Afghan. I recall his closest proxy was Tajik @ 8.5. You may be in a better position to comment on his likely ancestry based on this info, and comparing to your scores :)

Amerindian 0.05%
Ancestral_Altaic 5.33%
South_Central_Asian 28.96%
Arctic 0
South_Indian 15.95%
Australoid 0.26%
Austronesian 0
Caucasian 23.44%
Archaic_Human 0
East_African 0.07%
East_Siberian 1.78%
European_Early_Farmers 3.71%
Khoisan 0
Melano_Polynesian 0
Archaic_African 0.17%
Near_East 4.24%
North_African 0
Paleo_Siberian 0
African_Pygmy 0
South_East_Asian 3.53%
Subsaharian 0
Tungus-Altaic 6.06%
European_Hunters_Gatherers 6.45%


It could be HRP0410. He is listed as half Ghilzai Pashtun/half Panjshiri Tajik on the Harappa spreadsheet but I think he has Kabuli ancestry as well.

The Barnacle
04-04-2015, 06:33 PM
It could be HRP0410. He is listed as half Ghilzai Pashtun/half Panjshiri Tajik on the Harappa spreadsheet but I think he has Kabuli ancestry as well.

Kabuli as in from Kabul or Hindus in Afghanistan?

Rukha
04-04-2015, 07:57 PM
Kabuli as in from Kabul or Hindus in Afghanistan?

From Kabul, pretty sure you told me about him. And I was mistaken...the "Tajik" on Kurd's PCA is someone else.

The Barnacle
04-04-2015, 10:47 PM
From Kabul, pretty sure you told me about him. And I was mistaken...the "Tajik" on Kurd's PCA is someone else.

Where is the person from then?

Rukha
04-04-2015, 11:37 PM
Where is the person from then?

Not sure, could be of a mixed background...their East Eurasian and Near East + EEF scores are elevated compared to mine.

wmehar
09-21-2015, 09:42 PM
These were his scores. We are not sure what his ancestry is other than Afghan. I recall his closest proxy was Tajik @ 8.5. You may be in a better position to comment on his likely ancestry based on this info, and comparing to your scores :)

Amerindian 0.05%
Ancestral_Altaic 5.33%
South_Central_Asian 28.96%
Arctic 0
South_Indian 15.95%
Australoid 0.26%
Austronesian 0
Caucasian 23.44%
Archaic_Human 0
East_African 0.07%
East_Siberian 1.78%
European_Early_Farmers 3.71%
Khoisan 0
Melano_Polynesian 0
Archaic_African 0.17%
Near_East 4.24%
North_African 0
Paleo_Siberian 0
African_Pygmy 0
South_East_Asian 3.53%
Subsaharian 0
Tungus-Altaic 6.06%
European_Hunters_Gatherers 6.45%

He looks similar to mine, here they are:

Population
Amerindian 1.92%
Ancestral_Altaic 2.68%
South_Central_Asian 23.87%
Arctic -
South_Indian 15.00%
Australoid -
Austronesian -
Caucasian 27.21%
Archaic_Human -
East_African -
East_Siberian -
European_Early_Farmers 10.20%
Khoisan -
Melano_Polynesian 0.98%
Archaic_African 0.34%
Near_East 7.17%
North_African 3.98%
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 6.61%

Father is a Punjabi Arain, mother is a Greek-Cypriot.

okarinaofsteiner
05-11-2018, 11:55 PM
I found a PCA chart of 21 of the 23 MDLP K23b reference populations somewhere on Anthrogenica. Obviously it doesn't come close to describing the genetic variation among these ancestry components but it does a decent job of showing how distant Eurasian and Sub-Saharan ancestry components are from each other.

23096


This diagram that I also found somewhere on here shows the phylogenetic relationship between the 23 ancestry components. I think "Archaic Human" = other (probably non-Neanderthal, maybe Denisovan?) hominid ancestry, and "Archaic African" = Sub-Saharan ancestry that isn't accounted for in the other African components.

"ANI" = "South_Central_Asian", "ASI" = "South_Indian", "ANE" = "Ancestral Altaic", "WHG" = "European Hunter-Gatherers", "Altaic" = "Tungus_Altaic".

23097


Here are the ~620 reference populations plotted on the global PCA against 22 of the ancestry components. I arbitrarily assigned "Archaic_African" to somewhere near the other Sub-Saharan components so the African populations would cluster away from the Eurasian populations.

23098

okarinaofsteiner
05-11-2018, 11:58 PM
Global PCA plot without "Archaic_African"

https://i.imgur.com/tqIlPHj.png


Global PCA plot with "Archaic_African"
https://i.imgur.com/B9SVtiQ.png


Original global PCA plot of MDLP K23b ancestry components
https://i.imgur.com/BmZRlTQ.jpg

oz
05-13-2018, 05:21 AM
1 European_Hunters_Gatherers 40.2
2 Caucasian 35.33
3 European_Early_Farmers 12.91
4 South_Central_Asian 3.07
5 Near_East 2.27
6 Arctic 1.73
7 North_African 1.2
8 Ancestral_Altaic 0.94
9 Tungus-Altaic 0.88
10 Archaic_African 0.67
11 Paleo_Siberian 0.52
12 Melano_Polynesian 0.15
13 African_Pygmy 0.15

Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Croat_BH_ @ 4.510422
2 Bosnian_ @ 4.577154
3 Croat_ @ 5.402145
4 Serb_BH_ @ 5.457299
5 Hungarian_Budapest_ @ 8.079535
6 Czech_ @ 8.990827
7 Serb_Serbia_ @ 9.267186
8 Slovenian_ @ 9.410699
9 Montenegrian_ @ 9.453588
10 Hungarian_ @ 9.499269
11 Slovak_ @ 10.552804
12 Ukrainian_West_ @ 10.645851
13 Macedonian_ @ 11.084929
14 Austrian_ @ 12.277772
15 Bulgarian_ @ 12.567115
16 Moksha_ @ 13.824753
17 German_ @ 13.924696
18 Ukrainian_Center_ @ 14.240121
19 German_East_ @ 14.250154
20 Gagauz_ @ 14.724239

Using 2 populations approximation:
1 50% Gagauz_ +50% Ukrainian_East_ @ 2.781645


Using 3 populations approximation:
1 50% Bosnian_ +25% Crimean_Tatar_Coast_ +25% Lithuanian_ @ 2.479403


Using 4 populations approximation:
++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++
1 Belarusian_Russian_ + Crimean_Tatar_Coast_ + Croat_BH_ + Slovenian_ @ 2.418724
2 Crimean_Tatar_Coast_ + Croat_ + Croat_ + Lithuanian_ @ 2.419474
3 Crimean_Tatar_Coast_ + Croat_BH_ + Mixed_East_Slav_ + Slovenian_ @ 2.421702
4 Bosnian_ + Crimean_Tatar_Coast_ + Croat_ + Lithuanian_ @ 2.431870
5 Belarusian_ + Crimean_Tatar_Coast_ + Croat_BH_ + Slovenian_ @ 2.441254
6 Crimean_Tatar_Coast_ + Croat_BH_ + Pole_ + Slovenian_ @ 2.449817
7 Crimean_Tatar_Coast_ + Croat_ + Lithuanian_ + Serb_BH_ @ 2.453676
8 Belarusian_Russian_ + Crimean_Tatar_Coast_ + Croat_ + Slovenian_ @ 2.456619
9 Crimean_Tatar_Coast_ + Croat_ + Mixed_East_Slav_ + Slovenian_ @ 2.459720
10 Bosnian_ + Crimean_Tatar_Coast_ + Lithuanian_ + Serb_BH_ @ 2.462247
11 Crimean_Tatar_Coast_ + Croat_ + Croat_BH_ + Lithuanian_ @ 2.470763
12 Crimean_Tatar_Coast_ + Croat_BH_ + Lithuanian_ + Serb_BH_ @ 2.473248
13 Bosnian_ + Bosnian_ + Crimean_Tatar_Coast_ + Lithuanian_ @ 2.479403
14 Belarusian_Russian_ + Bosnian_ + Crimean_Tatar_Coast_ + Slovenian_ @ 2.482186
15 Belarusian_Russian_ + Crimean_Tatar_Coast_ + Croat_BH_ + Hungarian_ @ 2.484699
16 Bosnian_ + Crimean_Tatar_Coast_ + Mixed_East_Slav_ + Slovenian_ @ 2.486288
17 Crimean_Tatar_Coast_ + Croat_BH_ + Hungarian_ + Mixed_East_Slav_ @ 2.487007
18 Crimean_Tatar_Coast_ + Lithuanian_ + Serb_BH_ + Serb_BH_ @ 2.522061
19 Belarusian_South_ + Crimean_Tatar_Coast_ + Croat_BH_ + Croat_BH_ @ 2.523789
20 Belarusian_Russian_ + Crimean_Tatar_Coast_ + Serb_BH_ + Slovenian_ @ 2.535815

Dibran
05-14-2018, 08:44 PM
...........

Idk if you're still accepting samples given the thread was started in 2015. If so if you could include the following I would greatly appreciate it!

Me:
Admix Results (sorted):

# Population Percent
1 Caucasian 38.95
2 European_Early_Farmers 26.68
3 European_Hunters_Gatherers 21.62
4 Near_East 6.21
5 North_African 2.57
6 South_Central_Asian 1.31
7 Ancestral_Altaic 1.03
8 East_Siberian 0.91
9 Melano_Polynesian 0.41
10 Archaic_Human 0.31

Mother: Puka, Albania

Admix Results (sorted):

# Population Percent
1 Caucasian 36.61
2 European_Early_Farmers 25.43
3 European_Hunters_Gatherers 21.04
4 Near_East 7.71
5 North_African 3.71
6 South_Central_Asian 2.15
7 Ancestral_Altaic 1.34
8 Arctic 0.74
9 Subsaharian 0.72
10 Archaic_Human 0.56

Father: Diber Vogel, Albania

Admix Results (sorted):

# Population Percent
1 Caucasian 38.41
2 European_Early_Farmers 26.27
3 European_Hunters_Gatherers 20.82
4 Near_East 4.89
5 South_Central_Asian 2.93
6 North_African 2.42
7 South_Indian 1.18
8 Paleo_Siberian 0.59
9 Arctic 0.45
10 Ancestral_Altaic 0.44
11 African_Pygmy 0.38
12 Melano_Polynesian 0.37
13 East_Siberian 0.28
14 Khoisan 0.2
15 Australoid 0.18
16 Austronesian 0.16
17 Archaic_Human 0.04

JerryS.
05-14-2018, 09:36 PM
Population
Amerindian -
Ancestral_Altaic 5.28
South_Central_Asian 1.60
Arctic -
South_Indian -
Australoid 0.43
Austronesian -
Caucasian 30.36
Archaic_Human 0.17
East_African -
East_Siberian -
European_Early_Farmers 29.09
Khoisan -
Melano_Polynesian -
Archaic_African 0.16
Near_East 2.13
North_African 0.74
Paleo_Siberian 0.35
African_Pygmy 0.32
South_East_Asian -
Subsaharian 0.17
Tungus-Altaic -
European_Hunters_Gatherers 29.21


here are mine. I am mostly N.W. European but the population oracles seem to be Balkan shifted for my Mediterranean.

Jeep
05-14-2018, 10:34 PM
Amerindian 1.47 Pct
Ancestral_Altaic 4.17 Pct
South_Central_Asian 6.65 Pct
Arctic 0.27 Pct
South_Indian -
Australoid 0.32 Pct
Austronesian -
Caucasian 24.40 Pct
Archaic_Human 0.43 Pct
East_African -
East_Siberian -
European_Early_Farmers 22.33 Pct
Khoisan 0.21 Pct
Melano_Polynesian 0.25 Pct
Archaic_African 0.34 Pct
Near_East -
North_African 1.27 Pct
Paleo_Siberian -
African_Pygmy -
South_East_Asian -
Subsaharian 0.72 Pct
Tungus-Altaic -
European_Hunters_Gatherers 37.12 Pct

asm
05-15-2018, 01:23 AM
Just tested the accuracy of K23.Half Azeri(this side includes some North Caucasian mixture too) Half Ashkenazi.The result:

1 Caucasian 37.95
2 South_Central_Asian 15.71
3 Near_East 11.68
4 European_Early_Farmers 11.56
5 European_Hunters_Gatherers 10.71
6 North_African 3.78
7 Ancestral_Altaic 3.52
8 Tungus-Altaic 2.08
9 Arctic 0.77
10 Austronesian 0.69
11 East_Siberian 0.53
12 South_Indian 0.51
13 Melano_Polynesian 0.3
14 Subsaharian 0.21

Single Population Sharing:

# Population (source) Distance
1 Turk_Adana ( ) 4.6
2 Turk_Aydin ( ) 6.93
3 Turk_Balikesir ( ) 7.78
4 Turk ( ) 8.27
5 Turk_Istanbul ( ) 8.82
6 Turk_Kayseri ( ) 9.02
7 Iraqi_Chaldean ( ) 10.25
8 Syrian_Jew ( ) 10.29
9 Kurd_North ( ) 10.88
10 Azeri ( ) 11.2
11 Georgian_Jew ( ) 11.7
12 Lebanese ( ) 11.75
13 Crimean_Tatar_Mountain ( ) 11.92
14 Lebanese_Muslim ( ) 11.94
15 Assyrian_Iraqi ( ) 11.95
16 Iraqi_Mandean ( ) 12.25
17 Kurd_South ( ) 12.44
18 Azov_Greek ( ) 12.46
19 Iraki ( ) 12.71
20 Uzbekistani_Jew ( ) 12.78

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 88.8% Turk_Adana ( ) + 11.2% Dutch ( ) @ 1.86
2 89.9% Turk_Adana ( ) + 10.1% Swede ( ) @ 1.87
3 89.4% Turk_Adana ( ) + 10.6% Dane ( ) @ 1.9
4 87.7% Turk_Adana ( ) + 12.3% German-Volga ( ) @ 1.9
5 88.6% Turk_Adana ( ) + 11.4% Swede_Saami ( ) @ 1.95
6 90.3% Turk_Adana ( ) + 9.7% Norwegian_East ( ) @ 1.96
7 91.1% Turk_Adana ( ) + 8.9% Norwegian_West ( ) @ 1.96
8 89.7% Turk_Adana ( ) + 10.3% Belgian ( ) @ 1.97
9 89.6% Turk_Adana ( ) + 10.4% Frisian ( ) @ 1.97
10 88.7% Turk_Adana ( ) + 11.3% North_German ( ) @ 1.98
11 86.4% Turk_Adana ( ) + 13.6% Austrian ( ) @ 1.98
12 90.1% Turk_Adana ( ) + 9.9% English ( ) @ 2.01
13 91.2% Turk_Adana ( ) + 8.8% Icelandic ( ) @ 2.02
14 90.6% Turk_Adana ( ) + 9.4% North_European ( ) @ 2.05
15 89.9% Turk_Adana ( ) + 10.1% Irish ( ) @ 2.05
16 91.1% Turk_Adana ( ) + 8.9% British ( ) @ 2.06
17 90.9% Turk_Adana ( ) + 9.1% English_Cornwall_GBR ( ) @ 2.06
18 90.7% Turk_Adana ( ) + 9.3% English_Kent_GBR ( ) @ 2.08
19 91% Turk_Adana ( ) + 9% CEU ( ) @ 2.08
20 88.5% Turk_Adana ( ) + 11.5% South_German ( ) @ 2.08

?

PoxVoldius
05-23-2018, 03:43 AM
Here's mine:

Population
Amerindian 0.22
Ancestral_Altaic 4.49
South_Central_Asian 2.95
Arctic -
South_Indian -
Australoid 0.16
Austronesian -
Caucasian 22.68
Archaic_Human -
East_African -
East_Siberian -
European_Early_Farmers 24.25
Khoisan 0.49
Melano_Polynesian 0.79
Archaic_African -
Near_East -
North_African 2.45
Paleo_Siberian -
African_Pygmy -
South_East_Asian 1.09
Subsaharian 0.51
Tungus-Altaic -
European_Hunters_Gatherers 39.87

For reference, on my family tree, there's a lot of German (from all over Germany, Prussia, Silesia, and 19th century Russian Poland); a lot of British Isles (from all over); at least 2 Mayflower passengers; a great-grandmother whose known ancestors were all Czech; colonial era Dutch, French, Belgian, and Swiss; and a 4th great grandfather from Quebec (no info on his ancestry prior to Quebec); and several brick walls in North America.