PDA

View Full Version : My Gedmatch results



dowpa113
07-12-2017, 01:18 AM
Really have no idea, but here is goes. Thanks to Alhan, just did what he did. :)

I am Kurmanci-Kurd. Looks like other kurds right?

MDLP K23b Admixture Proportions


Amerindian 1.64
Ancestral_Altaic 3.80
South_Central_Asian 24.38
Arctic -
South_Indian 3.61
Australoid 0.33
Austronesian 1.19
Caucasian 40.92
Archaic_Human 0.22
East_African 0.45
East_Siberian -
European_Early_Farmers 2.75
Khoisan -
Melano_Polynesian 0.58
Archaic_African -
Near_East 12.27
North_African 3.38
Paleo_Siberian 1.03
African_Pygmy -
South_East_Asian 0.18
Subsaharian -
Tungus-Altaic -
European_Hunters_Gatherers 3.26

dowpa113
07-12-2017, 01:28 AM
Admix Results (sorted):

# Population Percent
1 Caucasian 40.92
2 South_Central_Asian 24.38
3 Near_East 12.27
4 Ancestral_Altaic 3.8
5 South_Indian 3.61
6 North_African 3.38
7 European_Hunters_Gatherers 3.26
8 European_Early_Farmers 2.75
9 Amerindian 1.64
10 Austronesian 1.19
11 Paleo_Siberian 1.03
12 Melano_Polynesian 0.58
13 East_African 0.45
14 Australoid 0.33
15 Archaic_Human 0.22
16 South_East_Asian 0.18

-



Single Population Sharing:

# Population (source) Distance
1 Baku_WGA ( ) 4.26
2 Azeri ( ) 5.08
3 Uzbek_Tashkent ( ) 5.39
4 Kurd_North ( ) 5.45
5 Kurd ( ) 5.53
6 Kurd_East ( ) 7.1
7 Uzbekistani_Jew ( ) 7.62
8 Kurd_South ( ) 8.87
9 Kurd_Jew ( ) 9.57
10 Iraqi_Mandean ( ) 9.73
11 Azeri_Dagestan ( ) 10.77
12 Iranian_Jew ( ) 11.22
13 Georgian_Jew ( ) 11.41
14 Turk_Adana ( ) 11.8
15 Tabassaran ( ) 12.21
16 Iraqi_Chaldean ( ) 12.28
17 Stalskoe_Kumyk ( ) 12.64
18 Lezgin ( ) 12.78
19 Lak ( ) 12.84
20 Avar ( ) 13.26

-

dowpa113
07-12-2017, 02:02 AM
Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance

1 95.4% Baku_WGA ( ) + 4.6% Moroccan ( ) @ 2.73

2 95.2% Baku_WGA ( ) + 4.8% Tunisian ( ) @ 2.79

3 96.5% Baku_WGA ( ) + 3.5% Saharawi ( ) @ 2.83

4 96.4% Baku_WGA ( ) + 3.6% Berber_WGA ( ) @ 2.83

5 95.4% Baku_WGA ( ) + 4.6% Shaigi_Sudan ( ) @ 2.84

6 92% Baku_WGA ( ) + 8% Egyptian_Tanta ( ) @ 2.85

7 95.7% Baku_WGA ( ) + 4.3% Algerian ( ) @ 2.89

8 96.2% Baku_WGA ( ) + 3.8% Morocco_South ( ) @ 2.9

9 86.9% Baku_WGA ( ) + 13.1% Muslim_Arabs_Israel ( ) @ 2.92

10 83.9% Baku_WGA ( ) + 16.1% Iraki ( ) @ 2.93

11 91.8% Baku_WGA ( ) + 8.2% Egyptian_Kafar_Sheikh ( ) @ 2.94

12 97% Baku_WGA ( ) + 3% Mozabite ( ) @ 2.94

13 91.7% Baku_WGA ( ) + 8.3% Egyptian_Mansoura ( ) @ 2.95

14 92% Baku_WGA ( ) + 8% Egyptian_Kairo ( ) @ 2.97

15 91.9% Baku_WGA ( ) + 8.1% Egyptian_Cairo ( ) @ 2.98

16 91% Baku_WGA ( ) + 9% BedouinA ( ) @ 2.99

17 92.6% Baku_WGA ( ) + 7.4% Egyptian_Kuwait ( ) @ 2.99

18 92.5% Baku_WGA ( ) + 7.5% Yemen ( ) @ 3.1

19 67.4% Iranian ( ) + 32.6% Adjara ( ) @ 3.1

20 89.6% Baku_WGA ( ) + 10.4% Jordanian ( ) @ 3.11

dowpa113
07-12-2017, 01:44 PM
Anyone? :)

Kurd
07-12-2017, 09:57 PM
Anyone? :)

It is best to view results of admixture tests in the context of membership in specific population clusters. This membership is not based on total shared drift histories, but rather based on alleles shared at a relatively few locations within the genome.

With regards to your results and results of Kurds in general using allele frequency based methods, we see a large membership in the Caucasian cluster followed by SC Asian cluster with the specific calculator you referenced. The ratios, SC Asian/ Caucasus will vary quite a bit based on the calculator, and will largely depend on which SC Asian and Caucasus populations have been elected to represent those particular clusters.

If N Caucasus samples are used as Caucasian references, then SC Asian / Caucasus for Kurds will increase, whereas if S Caucasus samples are used, the ratio will decrease.

By no means should you infer relatedness to a particular cluster based on the percentage you score in that cluster. Here is an example to illustrate:

Often we see N Indians score higher SC Asian than Kurds on various admixture tests, does that imply that N Indians are more related to SC Asians than Kurds are to SC Asians, or that they share more total genetic drift with them. The answer is no, and here is why.

You will notice that calculators that show such results have a W Asian component based on Caucasians and usually a S Indian/ S Asian component based on S Indian tribals. Let's say that such a calculator shows the following:

- Kurds score 50% W Asian and 25% SC Asian;
- N Indians score 35% S Indian and 35% or 45% SC Asian. The exact % is immaterial;

So at 1st glance it would appear that N Indians are more closely related to SC Asians than Kurds are to SC Asians, but the actual reason N Indians score higher SC Asian is not because they are more related to them, but because of the choice of the other calculator components, namely W Asian and S Indian.

Here the genetic distance Kurd to Caucasian is less than the genetic distance N Indian to S Indian tribal. The net effect is a higher percentage of Kurd admixture is allocated to the Caucasus cluster, which in turn leaves less to be allocated to the SC Asian cluster, whereas with N Indians a smaller allocation to the S Indian Tribal cluster occurs due to the greater distance N Indian <---> S Indian Tribal, thus leaving a larger portion to be allocated to the SC Asian cluster.

So, if we were to take the same calculator, and change the S Indian Tribal component to C or N Indian component, in other words make the component as close to N Indians as the Caucasus component is to Kurds, then Kurds will score higher SC Asian than N Indians, thus appearing to be more closely related to SC Asians, as most of the N Indian admixture will get picked up by the new C Or N Indian component, thus lowering their SC Asian score.

You can extrapolate the above to all components of an admixture calculator. The only way to tell who is closer to who is by one to one direct comparisons for shared drift with IBS, IBD, or statistical tools such as dstats

Kurd
07-12-2017, 10:10 PM
With regards to your other question, yes you fall within the normal variation for Kurmanji Kurds. Here are 2 Kurmanji Kurds from N Iraq from the Dohuk area


MDLP K23b Oracle Rev 2014 Sep 16

Admix Results (sorted):



#
Population
Percent


1
Caucasian
38.25


2
South_Central_Asian
26.20


3
Near_East
12.38


4
European_Early_Farmers
6.54


5
North_African
3.32


6
Ancestral_Altaic
3.20


7
South_Indian
2.50


8
European_Hunters_Gatherers
2.18


9
Tungus-Altaic
1.78


10
East_African
1.54




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

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

Least-squares method.

Using 1 population approximation:
1 Kurd_East @ 3.181808
2 Kurd @ 3.528385
3 Kurd_South @ 3.537720
4 Kurd_North @ 3.623128
5 Azeri @ 4.156168
6 Uzbek_Tashkent @ 6.682216
7 Baku_WGA @ 7.017886
8 Uzbekistani_Jew @ 8.340154
9 Iranian @ 9.417970
10 Turk_Adana @ 10.434705
11 Kurd_Jew @ 10.856094
12 Iraqi_Mandean @ 11.089017
13 Iraqi_Chaldean @ 12.139464
14 Iranian_Jew @ 12.375172
15 Turk_Kayseri @ 12.389807
16 Georgian_Jew @ 12.390733
17 Assyrian_Iraqi @ 12.664056
18 Azeri_Dagestan @ 13.565539
19 Jew_Tat @ 13.829996
20 Iraki @ 13.854867

Using 2 populations approximation:
1 50% Azeri +50% Kurd_South @ 2.539675


Using 3 populations approximation:
1 50% Azeri +25% Iranian +25% Kurd @ 2.427991


Using 4 populations approximation:
++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++
1 Iranian + Kurd + Kurd + Turk_Kayseri @ 2.218434
2 Iranian + Kurd + Kurd + Turk_Adana @ 2.289659
3 Azeri + Kurd + Kurd_South + Kurd_South @ 2.327182
4 Azeri + Kurd + Kurd_East + Kurd_South @ 2.344457
5 Azeri + Iranian + Kurd + Kurd_North @ 2.351383
6 Azeri + Azeri + Iranian + Kurd @ 2.427991
7 Brahui + Georgian_Svan + Samaritian + Turk_Kayseri @ 2.439063
8 Azeri + Kurd_East + Kurd_North + Kurd_South @ 2.457558
9 Abkhasian + Makrani + Samaritian + Turk_Kayseri @ 2.460124
10 Abkhasian + Makrani + Samaritian + Turk_Adana @ 2.466089




MDLP K23b Oracle Rev 2014 Sep 16

Admix Results (sorted):



#
Population
Percent


1
Caucasian
37.38


2
South_Central_Asian
26.92


3
Near_East
12.71


4
European_Early_Farmers
8.45


5
South_Indian
3.17


6
North_African
2.98


7
European_Hunters_Gatherers
1.75


8
Arctic
1.73




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

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

Least-squares method.

Using 1 population approximation:
1 Kurd_South @ 4.049366
2 Kurd_East @ 4.141598
3 Kurd_North @ 5.156777
4 Kurd @ 5.521064
5 Azeri @ 6.681501
6 Uzbek_Tashkent @ 8.288447
7 Iranian @ 9.028382
8 Baku_WGA @ 9.712027
9 Uzbekistani_Jew @ 9.762547
10 Turk_Adana @ 11.100921
11 Iraqi_Mandean @ 11.455456
12 Kurd_Jew @ 11.542463
13 Iraqi_Chaldean @ 12.470810
14 Assyrian_Iraqi @ 12.969651
15 Iranian_Jew @ 13.103626
16 Turk_Kayseri @ 13.159698
17 Georgian_Jew @ 13.308020
18 Iraki @ 14.699589
19 Jew_Tat @ 14.774176
20 Turk @ 15.009816

Using 2 populations approximation:
1 50% Kurd_East +50% Kurd_South @ 3.913028


Using 3 populations approximation:
1 50% Armenian +25% Brahui +25% Lebanese_Muslim @ 2.480702


Using 4 populations approximation:

1 Armenian + Brahui + Lebanese_Druze + Turk_Kayseri @ 2.148005
2 Brahui + Iraqi_Jew + Turk_Adana + Turk_Trabzon @ 2.248750
3 Armenian + Brahui + Lebanese_Christian + Turk_Kayseri @ 2.265373
4 Armenian_Yerevan + Brahui + Lebanese_Druze + Turk_Kayseri @ 2.284125
5 Armenian + Brahui + Cypriot + Jew_Tat @ 2.286771
6 Assyrian_Arzni + Balochi + Cypriot + Uzbek_Tashkent @ 2.289089
7 Armenian + Balochi + Iraqi_Jew + Turk_Kayseri @ 2.367000
8 Armenian + Cypriot + Makrani + Uzbek_Tashkent @ 2.374352
9 Armenian_Yerevan + Brahui + Cypriot + Jew_Tat @ 2.376543
10 Brahui + Cypriot + Iraqi_Chaldean + Kakheti @ 2.377272

dowpa113
07-12-2017, 11:43 PM
It is best to view results of admixture tests in the context of membership in specific population clusters. This membership is not based on total shared drift histories, but rather based on alleles shared at a relatively few locations within the genome.

With regards to your results and results of Kurds in general using allele frequency based methods, we see a large membership in the Caucasian cluster followed by SC Asian cluster with the specific calculator you referenced. The ratios, SC Asian/ Caucasus will vary quite a bit based on the calculator, and will largely depend on which SC Asian and Caucasus populations have been elected to represent those particular clusters.

If N Caucasus samples are used as Caucasian references, then SC Asian / Caucasus for Kurds will increase, whereas if S Caucasus samples are used, the ratio will decrease.

By no means should you infer relatedness to a particular cluster based on the percentage you score in that cluster. Here is an example to illustrate:

Often we see N Indians score higher SC Asian than Kurds on various admixture tests, does that imply that N Indians are more related to SC Asians than Kurds are to SC Asians, or that they share more total genetic drift with them. The answer is no, and here is why.

You will notice that calculators that show such results have a W Asian component based on Caucasians and usually a S Indian/ S Asian component based on S Indian tribals. Let's say that such a calculator shows the following:

- Kurds score 50% W Asian and 25% SC Asian;
- N Indians score 35% S Indian and 35% or 45% SC Asian. The exact % is immaterial;

So at 1st glance it would appear that N Indians are more closely related to SC Asians than Kurds are to SC Asians, but the actual reason N Indians score higher SC Asian is not because they are more related to them, but because of the choice of the other calculator components, namely W Asian and S Indian.

Here the genetic distance Kurd to Caucasian is less than the genetic distance N Indian to S Indian tribal. The net effect is a higher percentage of Kurd admixture is allocated to the Caucasus cluster, which in turn leaves less to be allocated to the SC Asian cluster, whereas with N Indians a smaller allocation to the S Indian Tribal cluster occurs due to the greater distance N Indian <---> S Indian Tribal, thus leaving a larger portion to be allocated to the SC Asian cluster.

So, if we were to take the same calculator, and change the S Indian Tribal component to C or N Indian component, in other words make the component as close to N Indians as the Caucasus component is to Kurds, then Kurds will score higher SC Asian than N Indians, thus appearing to be more closely related to SC Asians, as most of the N Indian admixture will get picked up by the new C Or N Indian component, thus lowering their SC Asian score.

You can extrapolate the above to all components of an admixture calculator. The only way to tell who is closer to who is by one to one direct comparisons for shared drift with IBS, IBD, or statistical tools such as dstats

Thanks alot, i really enjoy reading your stuff. :)

alhan
07-13-2017, 02:38 PM
With regards to your other question, yes you fall within the normal variation for Kurmanji Kurds. H

If we focus more on the Caucasian population cluster on Dowpa113's and mine
http://www.anthrogenica.com/showthread.php?11155-Understanding-my-Gedmatch-calculator-results


does the difference of around 3% from Dohuk's samples is just a variation or can we interpret this as a sign of real Caucasian (Cherkessian!) family history that we both separately have.

I would put all four results of MDLP K23b on the same table. But I was not properly able to manage it.

The second question is as we (Dowpa113 and myself) are searching for the confirmation of Cherkessian input into our gene pool, which calculator would be the best one in terms of reference populations.

Kurd
07-13-2017, 05:03 PM
If we focus more on the Caucasian population cluster on Dowpa113's and mine
http://www.anthrogenica.com/showthread.php?11155-Understanding-my-Gedmatch-calculator-results


does the difference of around 3% from Dohuk's samples is just a variation or can we interpret this as a sign of real Caucasian (Cherkessian!) family history that we both separately have.

I would put all four results of MDLP K23b on the same table. But I was not properly able to manage it.

The second question is as we (Dowpa113 and myself) are searching for the confirmation of Cherkessian input into our gene pool, which calculator would be the best one in terms of reference populations.

This would be impossible to ascertain with these types of admixture tests. I recommend an IBD test which I can possibly do for you guys when I return home in a week, but I would need your 23andme data for that. Contact me in 2 weeks and I will let you know if I have time. First, I owe John at Gedmatch an improved admixture calculator, and I need to wrap up an Iphone rare alleles admixture app.

It looks like you two are from Turkey.

dowpa113
07-13-2017, 08:50 PM
If we focus more on the Caucasian population cluster on Dowpa113's and mine
http://www.anthrogenica.com/showthread.php?11155-Understanding-my-Gedmatch-calculator-results


does the difference of around 3% from Dohuk's samples is just a variation or can we interpret this as a sign of real Caucasian (Cherkessian!) family history that we both separately have.

I would put all four results of MDLP K23b on the same table. But I was not properly able to manage it.

The second question is as we (Dowpa113 and myself) are searching for the confirmation of Cherkessian input into our gene pool, which calculator would be the best one in terms of reference populations.

Hey, nice to see you. :)

So we are Circassians now, i have heard that remarks. Alhan ur the only one that makes me to dig deeper.

Are we that now? I am 190cm, muscular and have a european face. My grand father speak fluent russian and had his russian hat on.



This would be impossible to ascertain with these types of admixture tests. I recommend an IBD test which I can possibly do for you guys when I return home in a week, but I would need your 23andme data for that. Contact me in 2 weeks and I will let you know if I have time. First, I owe John at Gedmatch an improved admixture calculator, and I need to wrap up an Iphone rare alleles admixture app.

It looks like you two are from Turkey.


Thanks for answering, i was okey with ur answer but alhan keeps pulling me up. :P

But whats with the Circassians remark? Its like when americans say the have native american ancestry. Why not armenian?

alhan
07-14-2017, 08:49 AM
This would be impossible to ascertain with these types of admixture tests. I recommend an IBD test which I can possibly do for you guys when I return home in a week, but I would need your 23andme data for that. Contact me in 2 weeks and I will let you know if I have time. First, I owe John at Gedmatch an improved admixture calculator, and I need to wrap up an Iphone rare alleles admixture app.

It looks like you two are from Turkey.

I understand you are very busy as you are on the side of doers unlike us who are just talkers.
But I would be happy to have you looking our data if you have time.
(I have Family Finder&YDNA12 instead of 23andme. )

Iam from Turkey and I think dowpa113 is also from Turkey as well. We do have similar stories; however not any known connection.

alhan
07-14-2017, 08:59 AM
Hey, nice to see you. :)

So we are Circassians now, i have heard that remarks. Alhan ur the only one that makes me to dig deeper.

Are we that now? I am 190cm, muscular and have a european face. My grand father speak fluent russian and had his russian hat on.

It's really not that easy to say if we have some kind of Circassian connection. But, if we do have that kind of history you are certainly more Cherkessian than me as your grand father speak fluent russian. My grand father was 1,70 and a ugly face (just like me) :angel:

Jokes aside, my possible Cherkessian connection is probably a lot deeper than yours. Should be at least 200-300 years probably more. On the other hand your grandfather speaking Russian is kind of strange because Adygea people are usually conservative and try to speak their own language.

By the way do you know, your haplogroup?

I made a comparison including your data in my thread
http://www.anthrogenica.com/showthread.php?10266-Ethnic-confirmation-of-Adyghe-origin&p=260400&viewfull=1#post260400

Kurd
07-15-2017, 09:43 AM
I understand you are very busy as you are on the side of doers unlike us who are just talkers.
But I would be happy to have you looking our data if you have time.
(I have Family Finder&YDNA12 instead of 23andme. )

Iam from Turkey and I think dowpa113 is also from Turkey as well. We do have similar stories; however not any known connection.

I normally don't take FTDNA, but I'll make an exception. The only thing is you will be in tests that don't require the higher marker overlaps, and I conduct tests time permitting. I'll be returning home next week. If the 2 of you want to be included in my dataset, you can send your raw data to [email protected] with info on what part of Turkey your 4 grand parents are from.

alhan
07-15-2017, 06:35 PM
I normally don't take FTDNA, but I'll make an exception. The only thing is you will be in tests that don't require the higher marker overlaps, and I conduct tests time permitting. I'll be returning home next week. If the 2 of you want to be included in my dataset, you can send your raw data to [email protected] with info on what part of Turkey your 4 grand parents are from.

Sounds very good to me.
It will be first thing Monday morning.
Thanks..

dowpa113
07-15-2017, 07:41 PM
It's really not that easy to say if we have some kind of Circassian connection. But, if we do have that kind of history you are certainly more Cherkessian than me as your grand father speak fluent russian. My grand father was 1,70 and a ugly face (just like me) :angel:

Jokes aside, my possible Cherkessian connection is probably a lot deeper than yours. Should be at least 200-300 years probably more. On the other hand your grandfather speaking Russian is kind of strange because Adygea people are usually conservative and try to speak their own language.

By the way do you know, your haplogroup?

I made a comparison including your data in my thread
http://www.anthrogenica.com/showthread.php?10266-Ethnic-confirmation-of-Adyghe-origin&p=260400&viewfull=1#post260400

True, i hope you find your ancestry do' :)

Na man, nobody is ugly. We are all equal. :)

Yea, its strange do'. Very nice, saw the turkic forum part.

My mother is U3B and father is L-M22, whats yours?

dowpa113
07-15-2017, 07:48 PM
The first is 23andme and the second is FamilyTreeDna, they say i am 90% Turkey/Armenia and 10% India.

https://anonimag.es/i/55540336.jpg

https://anonimag.es/i/34234f839.jpg

alhan
07-15-2017, 07:49 PM
True, i hope you find your ancestry do' :)

Na man, nobody is ugly. We are all equal. :)

Yea, its strange do'. Very nice, saw the turkic forum part.

My mother is U3B and father is L-M22, whats yours?

I had YDNA-12, resulted something I haven't been expecting R-M173.

I didn't have my fathers or my mother tested.

By the way I 'm sending my raw data to Kurd's e-mail with the info he asked, Are you?

dowpa113
07-15-2017, 07:54 PM
I had YDNA-12, resulted something I haven't been expecting R-M173.

I didn't have my fathers or my mother tested.

By the way I 'm sending my raw data to Kurd's e-mail with the info he asked, Are you?


Wait, is your mother YDNA-12 and father R-M173? Yes i am.

Edit, no your parents DNA in on the website like 23andme.

alhan
07-15-2017, 07:57 PM
Wait, is your mother YDNA-12 and father R-M173? Yes i am.

YDNA is the name of the test. Testing my paternal ancestry.

My haplogroup is R-M173.

No test for my mother or father.

dowpa113
07-15-2017, 08:05 PM
YDNA is the name of the test. Testing my paternal ancestry.

My haplogroup is R-M173.

No test for my mother or father.

Yea, the expensive one right. If you used 23andme then you will see your parents Haplogroup on "Ancestry Reports". Both your parents, it should be the same on Ancestry and sites like that. But you can of course test him also. But your parents Haplogroup tells you how they came out of Africa, if i am not wrong. It should be included with your RAW data when you buy it.

dowpa113
07-15-2017, 08:32 PM
My relatives are very interesting, most of them are in the USA. But its so mixed also. :)

https://anonimag.es/i/dhs216b0f6.jpg

alhan
07-16-2017, 05:40 PM
My relatives are very interesting, most of them are in the USA. But its so mixed also. :)

https://anonimag.es/i/dhs216b0f6.jpg

Very interesting. 23and me results are so different from FTDNA. I don't know what Im seeing.

dowpa113
07-20-2017, 04:46 PM
Very interesting. 23and me results are so different from FTDNA. I don't know what Im seeing.

Thats my relatives, their backround. Dont you have stuff like that? Would love to see it. :)

alhan
07-23-2017, 07:22 AM
Thats my relatives, their backround. Dont you have stuff like that? Would love to see it. :)

Yes, I have something similar. I will post it when I have access to my pc.

But having 90% Middle Eastern background and all DNA relatives from US and Europe is interesting.

My ydna relatives are also from a wide range, most from Eastern Europe, Sweden, Kyrgyzistan, India.

And I guess it's normal for R1a.

alhan
07-24-2017, 11:23 AM
This is what I have

17719

At Genetic Distance-1; compared to the country total. I am matching 12 out of all 84 Kyrgyz:behindsofa:
Similarly 3/39 in Bangladesh.

I have 9 perfect match from Norway, but it is just 0.3%.