PDA

View Full Version : Post your Yamnaya related K6 for graphs/PCAs



Kurd
04-22-2015, 10:54 PM
Post your Yamanya related K6 results for future analysis through graphs & PCAs. It would be interesting to compare with other members and references on the spreadsheet. Once I upload the PCAs you can compare and contrast your distances to other members and various spreadsheet reference populations with those from the K9 thread.

Please post your results in the same order as mine, to facilitate data entry into the spreadsheet.

As a heads up, I will continue uploading graphs into the K9 thread. English and Armenian are next.
Mine:

http://i.imgur.com/VQUDF5K.jpg
http://i.imgur.com/37kCZtW.jpg

Salkin
04-22-2015, 11:09 PM
Yamnaya_related 47.3481%
WHG_extra 3.8255%
ENA 0.9784%
Middle_Eastern 0.001%
Pre-Yamnaya 47.2667%
Sub-Saharan 0.5803%

Arbogan
04-22-2015, 11:09 PM
Yamnaya_related 0.193316
WHG_extra 0.012873
ENA 0.035195
Middle_Eastern 0.66386
Pre-Yamnaya 0.07946
Sub-Saharan 0.015295

Bulut
04-22-2015, 11:13 PM
K6

Yamnaya_related 21.0853%
WHG_extra 1.3608%
ENA 7.0521%
Middle_Eastern 50.9836%
Pre-Yamnaya 18.026%
Sub-Saharan 1.4922%

Sapporo
04-22-2015, 11:47 PM
I never got this done. Do I need to request it from David?

Kurd
04-22-2015, 11:49 PM
I never got this done. Do I need to request it from David?
If he did your k8, just shoot him an email

Sapporo
04-22-2015, 11:54 PM
If he did your k8, just shoot him an email

Thanks. :)

Rukha
04-23-2015, 01:05 AM
Yamnaya_related 36.4984%
WHG_extra 1.0715%
ENA 10.0112%
Middle_Eastern 50.4127%
Pre-Yamnaya 1E-005
Sub-Saharan 2.0052%

Helgenes50
04-23-2015, 02:20 AM
K6

Yamnaya_related 37.3425
WHG_extra 2.5659
ENA 0.001
Middle_Eastern 4.2168
Pre-Yamnaya 54.9468
Sub-Saharan 0.9271




Thanks

Jessie
04-23-2015, 02:26 AM
K6 results

Yamnaya_related 0.424978
WHG_extra 0.025084
ENA 0.008418
Middle_Eastern 0.048719
Pre-Yamnaya 0.492791
Sub-Saharan 0.00001

SwampThing27
04-23-2015, 03:50 AM
Yamnaya_related - 39.5%
WHG_extra - 2%
ENA - 2%
Middle_Eastern - 5.5%
Pre-Yamnaya - 50%
Sub-Saharan - 1%

CelticGerman
04-23-2015, 05:06 AM
My result:

Yamnaya_related 43.3645%
WHG_extra 2.6067%
ENA 0.001%
Middle_Eastern 2.7434%
Pre-Yamnaya 51.2834%
Sub-Saharan 0.001%

CelticGerman
04-23-2015, 05:22 AM
Here comes a French result, not related to me. You could call him JeanB. His known ancestry is entirely from Eastern France (Burgundy and Rhône-Alpes).

Yamnaya_related 29.8485
WHG_extra 2.0861
ENA 1.3678
Middle_Eastern 8.9004
Pre-Yamnaya 56.742
Sub-Saharan 1.0552

Jessie
04-23-2015, 05:44 AM
Just trying to understand this a bit better. Is the Pre-Yamnaya related to Neolithic ancestry in Europe?

Helgenes50
04-23-2015, 05:58 AM
[QUOTE=Jessie;80795]Just trying to understand this a bit better. Is the Pre-Yamnaya related to Neolithic ancestry in Europe?[/QUOTE

The Pre-Yamnaya in Western Europe probably is a mixture of Neolithics and WHGs, ie the MN Neolithics.

NK19191
04-23-2015, 12:10 PM
Yamnaya_related 0.18374
WHG_extra 0.009363
ENA 0.048309
Middle_Eastern 0.632109
Pre-Yamnaya 0.111056
Sub-Saharan 0.015423

Arbogan
04-23-2015, 12:14 PM
Yamnaya_related 0.18374
WHG_extra 0.009363
ENA 0.048309
Middle_Eastern 0.632109
Pre-Yamnaya 0.111056
Sub-Saharan 0.015423

I'm surprised you scored lower yamnaya than I did.

Anabasis
04-23-2015, 12:16 PM
Yamnaya_related 3,98160%
WHG_extra 1,78040%
ENA 0,00100%
Middle_Eastern 77,12270%
Pre-Yamnaya 17,11330%
Sub-Saharan 0,00100%

MfA
04-23-2015, 12:29 PM
KurdJew4633
Yamnaya_related 15.31%
WHG_extra 1.42%
ENA 2.74%
Middle_Eastern 65.59%
Pre-Yamnaya 14.30%
Sub-Saharan 0.64%

Kurd
04-23-2015, 12:40 PM
Yamnaya_related 3,98160%
WHG_extra 1,78040%
ENA 0,00100%
Middle_Eastern 77,12270%
Pre-Yamnaya 17,11330%
Sub-Saharan 0,00100%

Your data shows that there is large admixture variation across Turkey. You and Bulut appear to be on the 2 opposite ends of the spectrum. It is interesting that your ME component is substantially higher than that of any of the K6 spreadsheet Turkish samples (highest was 61%), and comparable to the that of Iraqi Jews :)

The Barnacle
04-23-2015, 12:42 PM
What is ENA again please?

Kurd
04-23-2015, 12:53 PM
What is ENA again please?

I believe it stands for Eastern Non African, or in other words East Eurasian. In k6 spreadsheet, it appears to peak in the Kazakh at about 45%, since there are no Chinese on the spreadsheet.

Anabasis
04-23-2015, 01:25 PM
Your data shows that there is large admixture variation across Turkey. You and Bulut appear to be on the 2 opposite ends of the spectrum. It is interesting that your ME component is substantially higher than that of any of the K6 spreadsheet Turkish samples (highest was 61%), and comparable to the that of Iraqi Jews :)

Whats interesting that i am very far away from Iraq, north east of turkey. I guess my ancestry have sticked in Kachkar mountains of blacksea for years and genetically isolated since ancient times. But there might be some migrations from mesopotamia to blacksea in Neolithic or pre Neolithic times.

Arbogan
04-23-2015, 01:32 PM
Whats interesting that i am very far away from Iraq, north east of turkey. I guess my ancestry have sticked in Kachkar mountains of blacksea for years and genetically isolated since ancient times. But there might be some migrations from mesopotamia to blacksea in Neolithic or pre Neolithic times.

More like that iraqi jew is acting as drain for all middle-east/caucasus like ancestry. People from trabzon seem to be most closely related to georgians. What are your K12B scores. You're some super caucasian-east-mediterranean mix, lol. Probably what north-eastern anatolia looked like before the arrival of indo-europeans.

Ignis90
04-23-2015, 02:12 PM
Mine:

Yamnaya_related 0.0046
WHG_extra 3.3926
ENA 3.9162
Middle_Eastern 41.1405
Pre-Yamnaya 34.7963
Sub-Saharan 16.8498



Fritalian, an Italian-like -mix friend:

Yamnaya_related 16.8032
WHG_extra 1.6548
ENA 0.001
Middle_Eastern 33.8241
Pre-Yamnaya 45.6928
Sub-Saharan 2.024

Anabasis
04-23-2015, 02:16 PM
More like that iraqi jew is acting as drain for all middle-east/caucasus like ancestry. People from trabzon seem to be most closely related to georgians. What are your K12B scores. You're some super caucasian-east-mediterranean mix, lol. Probably what north-eastern anatolia looked like before the arrival of indo-europeans.

I dont want to tak about out of the topic more. But to end the discussion about my scores here is my K12b Scores.

Western European 3.64%
Siberian -
East African -
West Central Asian 7.25%
South Asian -
West African 0.11%
Caucasus 54.88%
Finnish -
Mediterranean 21.48%
Southwest Asian 12.64%
North European -
East Asian -

Georgians i think they have more caucasian shifts then me. But I have more mediterrian and sw asian scores.

Kurd
04-23-2015, 02:20 PM
Mine:

Yamnaya_related 0.0046
WHG_extra 3.3926
ENA 3.9162
Middle_Eastern 41.1405
Pre-Yamnaya 34.7963
Sub-Saharan 16.8498



Fritalian, an Italian-like -mix friend:

Yamnaya_related 16.8032
WHG_extra 1.6548
ENA 0.001
Middle_Eastern 33.8241
Pre-Yamnaya 45.6928
Sub-Saharan 2.024


Do you have mixed ancestry. I need your ancestry for color coding on the graphs and charts. I am not able to guess it from the k6 spreadsheet as There does not appear to be any African groups on it and you have quiet a unique combo of SSA and ENA compared to the samples

Dr_McNinja
04-23-2015, 02:48 PM
HRP0393 Haryana Jatt (Beniwal)

Yamnaya_related 0.396573
WHG_extra 0.027646
ENA 0.143771
Middle_Eastern 0.410162
Pre-Yamnaya 0.00001
Sub-Saharan 0.021838

Ignis90
04-23-2015, 02:54 PM
Do you have mixed ancestry. I need your ancestry for color coding on the graphs and charts. I am not able to guess it from the k6 spreadsheet as There does not appear to be any African groups on it and you have quiet a unique combo of SSA and ENA compared to the samples

I forgot again, sorry. No I am not mixed, I am Berber (North African).

The ENA is not real and is very likely additional non-OOA (African) ancestry and also maybe some very basal Eurasian affinity as it often does in MENA people (to a much lesser extent here though).

The Barnacle
04-23-2015, 03:02 PM
HRP0393 Haryana Jatt (Beniwal)

Yamnaya_related 0.396573
WHG_extra 0.027646
ENA 0.143771
Middle_Eastern 0.410162
Pre-Yamnaya 0.00001
Sub-Saharan 0.021838

Can you post the results of the afghan pashtun HRP0370 please?

Dr_McNinja
04-23-2015, 03:32 PM
Can you post the results of the afghan pashtun HRP0370 please?

HRP0393 and HRP0341 were the only ones I sent to David for this one.

Speaking of which,

HRP0341 (Sapporo, Punjabi Jatt Sikh):

Yamnaya_related 0.395083
WHG_extra 0.009427
ENA 0.123066
Middle_Eastern 0.446615
Pre-Yamnaya 0.00001
Sub-Saharan 0.025798

MonkeyDLuffy
04-23-2015, 06:38 PM
Mine (punjabi tarkhan/ramgahia)

Yamnaya_relate 0.376751
WHG_extra 0.002609
ENA. 0.154832
Middle_Eastern 0.439909
Pre-Yamnaya 0.00001
Sub-Saharan 0.025889

randwulf
04-23-2015, 08:52 PM
Me:

Yamnaya_related 0.338062
WHG_extra 0.042703
ENA .007404
Middle_Eastern 0.121365
Pre-Yamnaya 0.481683
Sub-Saharan 0.008783

Father (German American):

Yamnaya_related 0.369217
WHG_extra 0.032378
ENA 0.011175
Middle_Eastern 0.079930
Pre-Yamnaya 0.491178
Sub-Saharan 0.016122


Mother (Colonial USA mix):

Yamnaya_related 0.359377
WHG_extra 0.047795
ENA 1E-005
Middle_Eastern 0.086189
Pre-Yamnaya 0.5015
Sub-Saharan 0.005129

CelticGerman
04-23-2015, 09:02 PM
randwulf, is your father's ancestry from the Palatinate or Hessian area? His Yamnaya percentage seems to be somehow in the middle between North Germans and Eastern French.

randwulf
04-23-2015, 09:13 PM
randwulf, is your father's ancestry from the Palatinate or Hessian area? His Yamnaya percentage seems to be somehow in the middle between North Germans and Eastern French.

I can be very specific - the towns are Aschaffenburg (Bavaria): 25%, Saarlouis (Saarland): 12.5%, Obergimpern (Baden): 25%, Darmstadt (Hesse) 25%, near Nuremberg (Bavaria): 12.5%. These are the direct immigrant town sources.

CelticGerman
04-23-2015, 10:00 PM
I can be very specific - the towns are Aschaffenburg (Bavaria): 25%, Saarlouis (Saarland): 12.5%, Obergimpern (Baden): 25%, Darmstadt (Hesse) 25%, near Nuremberg (Bavaria): 12.5%. These are the direct immigrant town sources.

Ok, that gives a geographic midpoint between Francfort and Heilbronn (with ancestry weighted according percentages you indicated). I think your Yamnaya percentage fits well with this position.

J Man
04-24-2015, 08:46 PM
Here are my results. :)

Yamnaya_related 0.403633
WHG_extra 0.03865
ENA 0.025522
Middle_Eastern 0.104178
Pre-Yamnaya 0.427023
Sub-Saharan 0.000994

Arbogan
04-24-2015, 09:00 PM
Here are my results. :)

Yamnaya_related 0.403633
WHG_extra 0.03865
ENA 0.025522
Middle_Eastern 0.104178
Pre-Yamnaya 0.427023
Sub-Saharan 0.000994
Interesting, do you have any first nation ancestry?

J Man
04-24-2015, 10:17 PM
Interesting, do you have any first nation ancestry?

No I do not. I am of 25% Italian, 25% Finnish and 50% Irish/British Isles ancestry.

BalkanKiwi
04-25-2015, 08:45 AM
Yamnaya_related 0.401338
WHG_extra 0.036672
ENA 0.009596
Middle_Eastern 0.063734
Pre-Yamnaya 0.488651
Sub-Saharan 0.00001

John Doe
04-25-2015, 09:12 AM
My results:

Yamnaya_related 0.148505
WHG_extra 0.007126
ENA 0.003979
Middle_Eastern 0.415615
Pre-Yamnaya 0.411474
Sub-Saharan 0.0133

Helgenes50
04-25-2015, 03:49 PM
Here comes a French result, not related to me. You could call him JeanB. His known ancestry is entirely from Eastern France (Burgundy and Rhône-Alpes).

Yamnaya_related 29.8485
WHG_extra 2.0861
ENA 1.3678
Middle_Eastern 8.9004
Pre-Yamnaya 56.742
Sub-Saharan 1.0552

Thanks for posting a french result.
the average of these is roughly of 32 %

In Normandy, ours are close to those of the English and the South Dutch
The difference, we are a little more neolithic, maybe more Danubian ( LBK)

France is a large and diverse country and genetically too.

Kurd
04-25-2015, 04:15 PM
Here is the community admixture graph. I have also included the averages of various populations from the K6 spreadsheet, as well as some ancient genomes.


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


EDIT: PCA should be uploaded later today.

CelticGerman
04-25-2015, 04:51 PM
I think the French "JeanB" is not included. Might be useful to add it too.

parasar
04-25-2015, 04:58 PM
Here is the community admixture graph. I have also included the averages of various populations from the K6 spreadsheet, as well as some ancient genomes.

http://i.imgur.com/7ICCRx9.jpg

EDIT: Just noticed that inadvertently Kenji was not included. Will update and repost in a couple of hours. Also, PCA should be uploaded later today.


A little surprised to see so little WHG and complete absence on pre-Yamnaya in northern South Asia. DLUFFY is at 0 WHG+pre-Yamnaya!

Kurd
04-25-2015, 05:07 PM
A little surprised to see so little WHG and complete absence on pre-Yamnaya in northern South Asia. DLUFFY is at 0 WHG+pre-Yamnaya!

I was also surprised. I am also still pondering ENA, and trying to account for S Asian admixture. Not sure whether Davidski's analysis or calculator may be a little off.

Helgenes50
04-25-2015, 05:52 PM
I was also surprised. I am also still pondering ENA. Not sure whether Davidski's analysis or calculator may be a little off.

Thanks for the french average

Most of the WHG in Western Europe is probably included in the pre-Yamnaya
or it wouldn't make sense. this one is a mixture of Farmers and of HGs

In my case, I get 45 % of WHG with K8 and only 2.5 % with K6

randwulf
04-25-2015, 06:43 PM
Can you post the results of the afghan pashtun HRP0370 please?


Thanks for the french average

Most of the WHG in Western Europe is probably included in the pre-Yamnaya
or it wouldn't make sense. this one is a mixture of Farmers and of HGs

In my case, I get 45 % of WHG with K8 and only 2.5 % with K6

Wasn't the category called "Extra WHG". That makes me think most WHG is in another category (pre-yamnaya) as you suggest. That must be happening with the south Asian admixture, too, maybe into ENA and middle east. Since the focus of this was to tease out similarity to yamnaya, it would be interesting to see Kurd's community admixture graph sorted from left to right primarily by the yamnaya component.

Kurd
04-25-2015, 07:06 PM
Wasn't the category called "Extra WHG". That makes me think most WHG is in another category (pre-yamnaya) as you suggest. That must be happening with the south Asian admixture, too, maybe into ENA and middle east. Since the focus of this was to tease out similarity to yamnaya, it would be interesting to see Kurd's community admixture graph sorted from left to right primarily by the yamnaya component.

ENA definitely does not contain WHG. ENA has some combination of E Eurasian & S Indian components, but I don't believe that it captures all of S Indian. For example, my ENA is 6.1%. If I back out my E Eurasian components, then the remainder is not sufficient to account for all of my S Asian/Indian. The same holds true for the other S Asian members. Yamnaya related can't be too Gedrosian/S Central Asian (Caucasus derived) heavy, because if it were, then we would not see such a high percentage of in Europeans, and relatively speaking low percentage in Iranians/Kurds.

I guess I will have to ponder and research a little more to be able to reconcile.

EDIT: it is also possible that, based on the references, this calculator may not be as informative for Asians as it is for Europeans

randwulf
04-25-2015, 07:38 PM
ENA definitely does not contain WHG. ENA has some combination of E Eurasian & S Indian components, but I don't believe that it captures all of S Indian. For example, my ENA is 6.1%. If I back out my E Eurasian components, then the remainder is not sufficient to account for all of my S Asian/Indian. The same holds true for the other S Asian members. Yamnaya related can't be too Gedrosian/S Central Asian (Caucasus derived) heavy, because if it were, then we would not see such a high percentage of in Europeans, and relatively speaking low percentage in Iranians/Kurds.

I guess I will have to ponder and research a little more to be able to reconcile.

EDIT: it is also possible that, based on the references, this calculator may not be as informative for Asians as it is for Europeans

I was not trying to say ENA contained WHG. I was agreeing that pre-yamnaya must have it. I was guessing that the south Asian stuff was being distributed in other categories like ENA and middle eastern (and some others?). Then, I was concluding the main purpose was to get to the Yamnaya value, so the other distinctions are of less importance to the task, so just move it out of the way, more or less. That led to the thought that the graph may be more informative sorted by Yamnaya, primarily. I hope I said it better this time.

Kurd
04-25-2015, 08:06 PM
I was not trying to say ENA contained WHG. I was agreeing that pre-yamnaya must have it. I was guessing that the south Asian stuff was being distributed in other categories like ENA and middle eastern (and some others?). Then, I was concluding the main purpose was to get to the Yamnaya value, so the other distinctions are of less importance to the task, so just move it out of the way, more or less. That led to the thought that the graph may be more informative sorted by Yamnaya, primarily. I hope I said it better this time.

Agreed, I will post by the end of the day

Chad Rohlfsen
04-25-2015, 08:11 PM
Pre-Yamnaya is like a Middle Neolithic marker. It has WHG.

Sein
04-25-2015, 08:32 PM
For what it's worth, this ADMIXTURE run is trying to show Neolithic + Bronze Age genetic ancestry. During the Neolithic, and during the Bronze Age, South Asians (as a genetic group) didn't exist. South Asian ADMIXTURE clusters are always complex mixtures of ancestry from the steppe+North Eurasia, ancestry from the ancient Near East, and very ancient ENA ancestry that is "autochthonous" to South Asia (and maybe also "indigenous" to the southern coastal edge of Southwestern Asia). In other words, modern South Asians aren't really an independent pole of genetic variation. Rather, they are a complex and rich mixture of some very divergent ancient populations, the product of different population expansions across different timescales, from the directions of Southwest Asia, Northern Eurasia, and from within South Asia itself. This complex spatio-temporal mixture, plus heavy drift that is structured by geography, create the South Asian genetic clusters we know and love (:biggrin1:).

So, this is why the calculator lacks a South Asian cluster (basically, since a South Asian cluster probably didn't exist at the timescale that we are querying).

The ENA percentages are very interesting, as this is a rare example of ADMIXTURE mirroring formal methods. In Haak et al., all Europeans had some ENA admixture, ranging from 1% to 10%. With f3 stats, most Europeans get a persistent negative signal involving the Dai. So this is in tune with those results.

For South Asians, ENA is basically standing in for all of the ASI ancestry, as well as for some the ENA shared by most West Eurasians in this run. We also have a Siberian/northeast Asian effect to consider, for some Pashtuns.

Krefter
04-25-2015, 08:36 PM
For what it's worth, this ADMIXTURE run is trying to show Neolithic + Bronze Age genetic ancestry. During the Neolithic, and during the Bronze Age, South Asians (as a genetic group) didn't exist. South Asian ADMIXTURE clusters are always complex mixtures of ancestry from the steppe+North Eurasia, ancestry from the ancient Near East, and very ancient ENA ancestry that is "autochthonous" to South Asia (and maybe also "indigenous" to the southern coastal edge of Southwestern Asia). In other words, modern South Asians aren't really an independent pole of genetic variation. Rather, they are a complex and rich mixture of some very divergent ancient populations, the product of different population expansions across different timescales, from the directions of Southwest Asia, Northern Eurasia, and from within South Asia itself. This complex spatio-temporal mixture, plus heavy drift that is structured by geography, create the South Asian genetic clusters we know and love (:biggrin1:).

So, this is why the calculator lacks a South Asian cluster (basically, since a South Asian cluster probably didn't exist at the timescale that we are querying).

The ENA percentages are very interesting, as this is a rare example of ADMIXTURE mirroring formal methods. In Haak et al., all Europeans had some ENA admixture, ranging from 1% to 10%. With f3 stats, most Europeans get a persistent negative signal involving the Dai. So this is in tune with those results.

For South Asians, ENA is basically standing in for all of the ASI ancestry, as well as for some the ENA shared by most West Eurasians in this run. We also have a Siberian/northeast Asian effect to consider, for some Pashtuns.

So, what makes South Asians differnt from West Eurasians is their own specific form of ENA? I thought what made South asians distinct was ancestry from a branch of Eurasians separate from ENA, WHG-ANE, and Basal Eurasian, connected to their unique mtDNA.

J Man
04-25-2015, 08:40 PM
The WHG_extra component is an interesting one in this test. It does look like most of our WHG ancestry is part of the Pre-Yamnaya component but what exactly then does the WHG_extra component represent? Why can't all of our WHG ancestry just be lumped into the Pre-Yamnaya component?

Chad Rohlfsen
04-25-2015, 08:42 PM
I'm working on some very interesting K runs which are unsupervised. David will post them onto Eurogenes later. K11 is very interesting, with a good hunter and Near East cluster that is peaked in a few EEF. Bedouins are separated from them by some South Asian, Central Asian, and SSA, in place of the WHG in the farmers. K12 is just about completed. Once I see some good clusters, I will do a few supervised runs, to get the best admixture results possible. If all looks good, perhaps a new calculator can be created.

J Man
04-25-2015, 08:45 PM
I'm working on some very interesting K runs which are unsupervised. David will post them onto Eurogenes later. K11 is very interesting, with a good hunter and Near East cluster that is peaked in a few EEF. Bedouins are separated from them by some South Asian, Central Asian, and SSA, in place of the WHG in the farmers. K12 is just about completed. Once I see some good clusters, I will do a few supervised runs, to get the best admixture results possible. If all looks good, perhaps a new calculator can be created.

That would be awesome if you and David could make a new calculator for us test junkies! :)

Sein
04-25-2015, 08:50 PM
So, what makes South Asians differnt from West Eurasians is their own specific form of ENA? I thought what made South asians distinct was ancestry from a branch of Eurasians separate from ENA, WHG-ANE, and Basal Eurasian, connected to their unique mtDNA.

This is a very interesting question.

One of the reasons for South Asian distinctiveness is genetic drift. South Asians are very drifted, compared to other populations. This South Asian-specific drift can make some admixed South Indian tribal groups seem unadmixed (close to 100% "South Asian" in some ADMIXTURE runs). Similar to what happens to Native Americans, but less extreme. Native Americans always turn out to be 100% "Native American", even though they are mixtures between North Eurasians and East Eurasians, and that too at variable levels.

But a big part of South Asian genetic distinctiveness is due to the fact that they are close to being only "Basal Eurasian" + ANE, with no WHG affinity. This is quite unique. In addition, South Asians have the most ANE out of all Eurasians.

But South Asians aren't really distinctive on global PCA plots. On any global PCA, the HGDP Pakistanis cluster close to populations from the Caucasus and Russia, but to their east (the eastward shift is due to both ANE and ENA).

Chad Rohlfsen
04-25-2015, 08:50 PM
That would be awesome if you and David could make a new calculator for us test junkies! :)

That'll be up to David. I can't make any promises, or speak on his behalf. I haven't learned how to do them yet. He's the only one with the marbles in that area. I'm still a rookie.

Chad Rohlfsen
04-25-2015, 08:54 PM
Using tribals will help to flush out some stuff.

Also, in the K10 on Eurogenes, a component that peaks in Yamnaya showed up (65%), separate from that in SC Asian cluster that peaked in Gujarati D and Kalash. SC Asians also had some of this . It has interesting implications, considering the Euro HG in Central and SC Asian pops. Both of which made up almost all of Yamnaya in that run.

Chad Rohlfsen
04-25-2015, 08:57 PM
The WHG_extra component is an interesting one in this test. It does look like most of our WHG ancestry is part of the Pre-Yamnaya component but what exactly then does the WHG_extra component represent? Why can't all of our WHG ancestry just be lumped into the Pre-Yamnaya component?

It's stuff in excess of the cluster created with those Neolithic samples, and outside of what Yamnaya samples offer.

Chad Rohlfsen
04-25-2015, 09:03 PM
For what it's worth. Oetzi peaked in the pre-Yamnaya component. He comes up 20% hunter in excess of KO2, in the K11. Not much more hunter than Stuttgart though. Still about 10-15% short of those like Esperstedt or Spain MN. Most early farmer samples are about 6-10% more Near Eastern than Oetzi. KO2 is pretty extreme. These extreme Near Eastern ones don't have a lot of markers, so don't get too excited. I can try them as the supervised farmers, to see if it's bogus or not.

Kurd
04-25-2015, 09:06 PM
For what it's worth, this ADMIXTURE run is trying to show Neolithic + Bronze Age genetic ancestry. During the Neolithic, and during the Bronze Age, South Asians (as a genetic group) didn't exist. South Asian ADMIXTURE clusters are always complex mixtures of ancestry from the steppe+North Eurasia, ancestry from the ancient Near East, and very ancient ENA ancestry that is "autochthonous" to South Asia (and maybe also "indigenous" to the southern coastal edge of Southwestern Asia). In other words, modern South Asians aren't really an independent pole of genetic variation. Rather, they are a complex and rich mixture of some very divergent ancient populations, the product of different population expansions across different timescales, from the directions of Southwest Asia, Northern Eurasia, and from within South Asia itself. This complex spatio-temporal mixture, plus heavy drift that is structured by geography, create the South Asian genetic clusters we know and love (:biggrin1:).

So, this is why the calculator lacks a South Asian cluster (basically, since a South Asian cluster probably didn't exist at the timescale that we are querying).

The ENA percentages are very interesting, as this is a rare example of ADMIXTURE mirroring formal methods. In Haak et al., all Europeans had some ENA admixture, ranging from 1% to 10%. With f3 stats, most Europeans get a persistent negative signal involving the Dai. So this is in tune with those results.

For South Asians, ENA is basically standing in for all of the ASI ancestry, as well as for some the ENA shared by most West Eurasians in this run. We also have a Siberian/northeast Asian effect to consider, for some Pashtuns.

Thanks Sein. I always find your posts quite informative.



For South Asians, ENA is basically standing in for all of the ASI ancestry, as well as for some the ENA shared by most West Eurasians in this run. We also have a Siberian/northeast Asian effect to consider, for some Pashtuns

That was my conclusion as well, although mine seemed too low to contain E Eurasian as well as any ASI type admixture, but I guess this is not the main emphasis of this calculator

Krefter
04-25-2015, 09:11 PM
For what it's worth. Oetzi peaked in the pre-Yamnaya component. He comes up 20% hunter in excess of KO2, in the K11. Not much more hunter than Stuttgart though. Still about 10-15% short of those like Esperstedt or Spain MN. Most early farmer samples are about 6-10% more Near Eastern than Oetzi. KO2 is pretty extreme. These extreme Near Eastern ones don't have a lot of markers, so don't get too excited. I can try them as the supervised farmers, to see if it's bogus or not.

Is Stuttgart the most Near eastern of early farmers?

Here's a random thought: Middle Neolithic and Copper age Italians may have had more ENF than others Euros, considering how much modern Sardinians do and what Otzei had. Making Italians part Otzei and part BR1 or LN/BA(representing Italic and other IE newcomers) might be realistic.

Italy may have been in its own world genetically during Pre-History. Pre-Historic genomes from mainland Europe might not be representative of what was happening in Italy during the same time.

Chad Rohlfsen
04-25-2015, 09:22 PM
K02 was tops at 87%, followed by Starcevo, LBKT, and a couple German LBK at about 80%.

K11 results are matching Dstats for hunters pretty well too!

Krefter
04-25-2015, 09:30 PM
But a big part of South Asian genetic distinctiveness is due to the fact that they are close to being only "Basal Eurasian" + ANE, with no WHG affinity. This is quite unique. In addition, South Asians have the most ANE out of all Eurasians.

No WHG affinity?! Would this also mean West Asians are mostly Basal Eurasian+ANE, and that East Asians are most distant from West Asians?

Coldmountains
04-25-2015, 09:32 PM
So, what makes South Asians differnt from West Eurasians is their own specific form of ENA? I thought what made South asians distinct was ancestry from a branch of Eurasians separate from ENA, WHG-ANE, and Basal Eurasian, connected to their unique mtDNA.
What is actually the relationship between ANE and WHG ? It looks for me rather that this two components are very different and ANE was probably created somewhere in Central Asia or Siberia far away from the populations who contributed to WHG.

Sein
04-25-2015, 09:37 PM
No WHG affinity?! Would this also mean West Asians are mostly Basal Eurasian+ANE, and that East Asians are most distant from West Asians?

Interestingly, there is some WHG affinity in West Asia, but it's shifted towards the Mediterranean basin, rather than Northern Europe.

Also, some South Asians, like Jatts, do display a WHG affinity. But unlike West Asians, their WHG affinity is towards Northern Europe, rather than the Mediterranean basin.

Sein
04-25-2015, 09:46 PM
What is actually the relationship between ANE and WHG ? It looks for me rather that this two components are very different and ANE was probably created somewhere in Central Asia or Siberia far away from the populations who contributed to WHG.

The Haak et al. paper proposes a very complex picture of WHG-ANE relationships. Based on their best fitting trees, EHG and ANE are sister clades, very closely related. But WHG is a mixture between the ANE/EHG clade and some sort of population basal to the ANE/EHG clade, probably K14. The special relationship between EHG and WHG holds because WHG is EHG + basal non-ENA Eurasian (probably K14, in my view).

This is demonstrated by these qpAdm fits. Using MA1 (ANE reference), David could fit the HGDP Pashtuns as 50.1% Near Eastern + 35.3% ANE + 14.6% ENA. Using the EHG samples, he could fit the HGDP Pashtuns as 48.8% Near Eastern + 37.2% EHG + 14.1% ENA. Interestingly, the EHG fit is much worse, so when it comes to the ANE/EHG clade, Pashtuns are probably ANE-admixed, rather than EHG-admixed.

Dr_McNinja
04-25-2015, 09:52 PM
Depending on how much WHG went into Yamnaya, some ENA also went into Yamnaya. Like Sapporo's ENA dropped almost 2% more than the Haryana Jatt's (Beniwal) even though they're usually around the same in ASE in other calculators. Because more of Sapporo's WHG got captured by Yamnaya, so more of the ASE/ENA also did as well. Thus he winds up at a high 39%, almost the same as the Haryana individual's 39%, even though they had leftover WHG.

For example:

Punjabi Tarkhan/Ramgarhia (Duffy):

West Eurasia K8 WHG: 2.72%
Yamnaya K6 WHG_extra: 0.26%

West Eurasia K8 ASE: 17.36% (+2.84% East_Eurasian, +2.98% Oceanian)
Yamnaya K6 ENA: 15.48%

That's a drop of around 2%.

Haryana Jatt:

West Eurasia K8 WHG: 7.41%
Yamnaya K6 WHG_extra: 2.76%

West Eurasia K8 ASE: 15.20% (+1.28% East_Eurasian, +2.47% Oceanian)
Yamnaya K6 ENA: 14.38%

That's around 1%.

Sapporo (Punjabi Jatt Sikh):

West Eurasia K8 WHG: 1.95%
Yamnaya K6 WHG_extra: 0.94%

West Eurasia K8 ASE: 15.51% (+1.81% East_Eurasian, +1.72% Oceanian)
Yamnaya K6 ENA: 12.31%

That's over a 3% drop.

Kenji:

West Eurasia K8 WHG: 0%
Yamnaya K6 WHG_extra: 1.25%

West Eurasia K8 ASE: 18.22% (+4.15% East_Eurasian, +4.38% Oceanian)
Yamnaya K6 ENA: 17.94%

No drop, he gained WHG that wasn't picked up in the K8. His Yamnaya was basically his ANE plus some East Eurasian. His K8 Oceanian went into Yamnaya, WHG_extra, Middle East, SSA, and Pygmy.

Sapporo, like many of the other Punjabi Jatt Sikhs, also has some heavier Uralic-type affinity in his East Eurasian from other calculators, which probably attracted more of it into Yamnaya.

That's why it's not the best calculator for South Asians. A Steppe component should be tracking WHG in South Asia.

Krefter
04-25-2015, 09:55 PM
What is actually the relationship between ANE and WHG ? It looks for me rather that this two components are very different and ANE was probably created somewhere in Central Asia or Siberia far away from the populations who contributed to WHG.

I don't know much about their relationship, you should ask Davidski at the Eurogenes blog. From recent Ancient DNA papers they've shown WHG and ANE form a clade as opposed to "East Asian".

The reason native Americans are closer to Loschbour than some Middle Eastern are, is because ANE and WHG are closely related.

I've heard that MA-1's culture was closely related to cultures that existed in Europe at the same time. K-14 was apart of this culture, but lived over 10,00 years before MA-1, and Davidski says he looks ancestral to ANE and WHG.

Dr_McNinja
04-25-2015, 10:03 PM
I think there's an as-yet undiscovered archaic component in South Asia, maybe peaking in the Himalayas, that acts like chameleon admixture. It looks like Oceanian in K8 which we all got significant percentages of, then disperses into everything else in other calculators.

Kurd
04-25-2015, 10:21 PM
Here is the community admixture graph sorted by Yamnaya score. I will have the PCA posted by the end of the day.

http://i.imgur.com/VPzn00S.jpg

randwulf
04-25-2015, 10:36 PM
Depending on how much WHG went into Yamnaya, some ENA also went into Yamnaya. Like Sapporo's ENA dropped almost 2% more than the Haryana Jatt's (Beniwal) even though they're usually around the same in ASE in other calculators. Because more of Sapporo's WHG got captured by Yamnaya, so more of the ASE/ENA also did as well. Thus he winds up at a high 39%, almost the same as the Haryana individual's 39%, even though they had leftover WHG.

My father usually shows 3-5% Eastern type admixture (add Siberian, east Asian, oceanian, etc.) type numbers into an otherwise west German admixture. In K6, this seems to show as ENA 1.1% plus SSA 1.6% plus maybe some in the yamnaya component. I show about one half his numbers in most tests on these and the same happens for k6.

Arbogan
04-25-2015, 10:39 PM
I think this is one of the better tests.

Pre-yamnaya: Neolithic Basal Eurasian+ Mediterranean WHG
Middle-eastern: Caucasus(basal Eurasian+ ANE)+ middle-eastern(Basal Eurasian) like Neolithic ancestry, non yamnaya related west-Asian/gedrosia(ANE+basal Eurasian+ASI).
SSA: Sub-Saharan(which nicely corresponds to African alleles in west-Asians, seems to give weird signals in south-central+south-asians)
Yamnaya: WHG+ANE+ whatever you want to call this Neolithic/west-Asian(or gedrosia) signal in yamnaya
ENA: ASI or East-Eurasian depending on admixture
WHG: hunter gather ancestry that doesn't fit in with other groups.

even though it's noisy, I think it gives pretty solid results.

newtoboard
04-25-2015, 10:44 PM
I don't know much about their relationship, you should ask Davidski at the Eurogenes blog. From recent Ancient DNA papers they've shown WHG and ANE form a clade as opposed to "East Asian".

The reason native Americans are closer to Loschbour than some Middle Eastern are, is because ANE and WHG are closely related.

I've heard that MA-1's culture was closely related to cultures that existed in Europe at the same time. K-14 was apart of this culture, but lived over 10,00 years before MA-1, and Davidski says he looks ancestral to ANE and WHG.

Y P almost certainly came from SE Asia. Disagreeing with that requires a lot of special pleading and trying to convince people of a lot of ridiculous scenarios. So no way Malta's ancestors originated in Europe. Not to mention R1a and R1b probably both originated in Asia not Europe and Jean's (and others) dating of EHG to about 8000 years(???) is much more likely.

Krefter
04-25-2015, 11:25 PM
Y P almost certainly came from SE Asia. Disagreeing with that requires a lot of special pleading and trying to convince people of a lot of ridiculous scenarios. So no way Malta's ancestors originated in Europe. Not to mention R1a and R1b probably both originated in Asia not Europe and Jean's (and others) dating of EHG to about 8000 years(???) is much more likely.

MA-1 had close cultural relatives in Europe. That's what knowledgeable people in archaeology say. Also, we know his cultural relatives in Europe were closely related to him, because we have K-14's genome and mtDNA U in Czech Republic that's a few thousand years older than MA-1. I never said MA-1 came from Europe, just that he had relatives in Europe, some who lived many thousands of years before he did.

R could be from SE Asia. That doesn't change MA-1's relation to K-14 and WHG. A distant origin somewhere far away in southern or eastern Asia makes sense considering ANE-affinity in South Asia. West Eurasian mtDNA, including U, though strictly points to a western origin, far away from East Asia.

Kurd
04-25-2015, 11:32 PM
Here is the PCA. About 65% of the variation is captured by PC1/PC2. I have also posted the eigenvectors so that you have an idea of the component axes. The eigenvalues will give you an idea of the variation that is captured (last graphic)


http://i.imgur.com/L45aqS2.jpg
http://i.imgur.com/le3wr3M.jpg
http://i.imgur.com/Cby66tx.jpg

newtoboard
04-25-2015, 11:32 PM
MA-1 had close cultural relatives in Europe. That's what knowledgeable people in archaeology say. Also, we know his cultural relatives in Europe were closely related to him, because we have K-14's genome and mtDNA U in Czech Republic that's a few thousand years older than MA-1. I never said MA-1 came from Europe, just that he had relatives in Europe, some who lived many thousands of years before he did.

R could be from SE Asia. That doesn't change MA-1's relation to K-14 and WHG. A distant origin somewhere far away in southern or eastern Asia makes sense considering ANE-affinity in South Asia. West Eurasian mtDNA, including U, though strictly points to a western origin, far away from East Asia.

He probably had more relatives and ancestors in Asia than Europe. We have an archeologist on this site who seems to think that his culture originated somewhere in Iran and Uzbekistan. So western origin doesn't mean Europe. And it should really be specified as SW as people might get confused. Especially since mt U could have originated at the Asian crossroads like Jean suggested. And other than U4/U5 most U clades are of Central -West Asian origin (U1, U2, U3, U6, U7, U8/U9).

Sein
04-25-2015, 11:35 PM
This is the best tree in Haak et al.

4428

The good thing is that it also models ASI, by including the Onge, and thus also accounts for ENA.

As is evident, ANE and EHG are very closely related (they are members of sister clades). Loschbour (WHG) is 11% EHG/ANE, and 89% of something quite far up the tree (but after the split with ENA).

Interestingly, K14 is much more closely related to WHG, closer than it is to EHG and ANE. So, that 89% divergent ancestry in WHG is probably from K14-like populations.

newtoboard
04-25-2015, 11:37 PM
Here is the PCA. About 65% of the variation is captured by PC1/PC2. I have also posted the eigenvectors so that you have an idea of the component axes. The eigenvalues will give you an idea of the variation that is captured (last graphic)


http://i.imgur.com/L45aqS2.jpg
http://i.imgur.com/le3wr3M.jpg
http://i.imgur.com/Cby66tx.jpg

Is it fair to say that Central Asian Scythians would have clustered between Yamnaya and Tajiks? And European Scythians between Yamnaya and Corded Ware?

If the Corded Ware (Abashevo) origin of Indo-Iranians is true this would suggest the amount of IE ancestry in Asia is close to zero given that Russian corded ware will probably look like Yamnaya with more of the WHG component and Indo-Iranians speakers are already getting very low WHG scores.

Krefter
04-25-2015, 11:44 PM
He probably had more relatives and ancestors in Asia than Europe. We have an archeologist on this site who seems to think that his culture originated somewhere in Iran and Uzbekistan. So western origin doesn't mean Europe. And it should really be specified as SW as people might get confused. Especially since mt U could have originated at the Asian crossroads like Jean suggested. And other than U4/U5 most U clades are of Central -West Asian origin (U1, U2, U3, U6, U7, U8/U9).

I never said he was from Europe or that's where his closest relatives were, I said he had close relatives there. He had close relatives in Asia to. I agree mtDNA U looks like a very old lineage from Central-West Asia. U6, U2, U7, U3, U1, U9, and U6 have clear origins there.

Kurd
04-25-2015, 11:45 PM
Here is the PCA. About 65% of the variation is captured by PC1/PC2. I have also posted the eigenvectors so that you have an idea of the component axes. The eigenvalues will give you an idea of the variation that is captured (last graphic)


http://i.imgur.com/L45aqS2.jpg
http://i.imgur.com/le3wr3M.jpg
http://i.imgur.com/Cby66tx.jpg

A review of the eigenvectors shows the following:

- Expectedly WHG & Pre Yamnaya are positively correlated. An increase/decrease in one results in a proportional increase/decrease in the other
- Expectedly ME & SSA are also positively correlated, with a lesser degree than the above
- ENA & Pre Yamnaya/WHG are negatively correlated. An increase in one is accompanied by a proportional decrease in the other
- Yamnaya & ENA are uncorrelated. An increase/decrease in one does not affect the other

Dr_McNinja
04-26-2015, 02:17 AM
Kurd, could you run the PCA again and put me on it when you get time? I came out a bit different than the other South Asians. My SSA is pretty high and almost none of my K8 WHG went anywhere.

My results:

Yamnaya_related 0.347662
WHG_extra 0.015382
ENA 0.167452
Middle_Eastern 0.428361
Pre-Yamnaya 0.00001
Sub-Saharan 0.041132

SSA seems like some kind of basal or SW-Asian-like Eurasian here, but it didn't go into Pre-Yamnaya (I got some EEF in K9). Just came out as SSA. I think some of it is definitely from Oceanian in K8.

I made a little chart for a few of the South/SC Asians showing the differences:

https://docs.google.com/spreadsheet/ccc?key=0AuXBmvmgdkfVdFMtRHVlZDBuQ3lMcjhxMDE4V3JoY lE#gid=58

Kurd
04-26-2015, 02:45 AM
Kurd, could you run the PCA again and put me on it when you get time? I came out a bit different than the other South Asians. My SSA is pretty high and almost none of my K8 WHG went anywhere.

My results:

Yamnaya_related 0.347662
WHG_extra 0.015382
ENA 0.167452
Middle_Eastern 0.428361
Pre-Yamnaya 0.00001
Sub-Saharan 0.041132

SSA seems like some kind of basal or SW-Asian-like Eurasian here, but it didn't go into Pre-Yamnaya (I got some EEF in K9). Just came out as SSA. I think some of it is definitely from Oceanian in K8.

I made a little chart for a few of the South/SC Asians showing the differences:

https://docs.google.com/spreadsheet/ccc?key=0AuXBmvmgdkfVdFMtRHVlZDBuQ3lMcjhxMDE4V3JoY lE#gid=58

Here you go: I am wondering if the lower SNP coverage for this calculator (possibly 50% of K9?) is having an effect. Generally, the lower the number of markers sampled, the lower the accuracy, and the greater the variation in the results between 2 indivduals of the same ethnic background. I would think that more markers would be needed to capture all of the regions on the genome associated with the allelic frequencies of all 6 components?.

http://i.imgur.com/2fYdQMa.jpg

Ignis90
04-26-2015, 12:27 PM
[SIZE=3]
http://i.imgur.com/2fYdQMa.jpg

Great!

Can you neutralize the effect of ENA and SSA in order to see where South Asians (and I) end up?

Krefter
04-26-2015, 01:07 PM
Kurd can you put me in the PCA? Here are my results.

Yamna_related: 36.5151
Pre_Yamna: 46.883
Middle_Eastern: 11.3951
WHG_extra: 2.436
ENA: 1.7989
Sub Saharan: 0.9719

John Doe
04-26-2015, 01:11 PM
Here is the PCA. About 65% of the variation is captured by PC1/PC2. I have also posted the eigenvectors so that you have an idea of the component axes. The eigenvalues will give you an idea of the variation that is captured (last graphic)


http://i.imgur.com/L45aqS2.jpg
http://i.imgur.com/le3wr3M.jpg
http://i.imgur.com/Cby66tx.jpg

I seem to be a little closer to Europe than other AJs, I know that I'm 1/16th or so Scandinavian (Danish), but is that enough to contribute to that?

Chad Rohlfsen
04-26-2015, 03:12 PM
......

Kurd
04-26-2015, 03:33 PM
I seem to be a little closer to Europe than other AJs, I know that I'm 1/16th or so Scandinavian (Danish), but is that enough to contribute to that?


There are a couple of AJs close to you, but the majority seem more ME shifted. You may want to look at the k6 spreadsheet to compare with them. Not sure what region those AJs are from

Kurd
04-26-2015, 03:34 PM
Kurd can you put me in the PCA? Here are my results.

Yamna_related: 36.5151
Pre_Yamna: 46.883
Middle_Eastern: 11.3951
WHG_extra: 2.436
ENA: 1.7989
Sub Saharan: 0.9719

Ok, will do next time I do a run

Kurd
04-26-2015, 03:35 PM
Great!

Can you neutralize the effect of ENA and SSA in order to see where South Asians (and I) end up?

What will we learn by doing this?

Ignis90
04-26-2015, 03:47 PM
What will we learn by doing this?

It's just a way to focus more on "Western Eurasia" 's diversity by eliminating signals of ancestry that aren't shared by at least the majority of the populations in presence. Basically, I'd like to see if South Asians overlap with some West/Central Asians in the strictly "West Eurasian" context. And also where I (Berber) stand vis-à-vis the farmer Stuttgart.

Chad Rohlfsen
04-26-2015, 04:12 PM
It's just a way to focus more on "Western Eurasia" 's diversity by eliminating signals of ancestry that aren't shared by at least the majority of the populations in presence. Basically, I'd like to see if South Asians overlap with some West/Central Asians in the strictly "West Eurasian" context. And also where I (Berber) stand vis-à-vis the farmer Stuttgart.

The Paniya and Austro-asiatic ones really don't share much with them, compared to groups like the Chamar, Hakkipikki, and Dusadh.

Kurd
04-26-2015, 04:27 PM
It's just a way to focus more on "Western Eurasia" 's diversity by eliminating signals of ancestry that aren't shared by at least the majority of the populations in presence. Basically, I'd like to see if South Asians overlap with some West/Central Asians in the strictly "West Eurasian" context. And also where I (Berber) stand vis-à-vis the farmer Stuttgart.

I presume what you mean is you would like to go back in time, and see what the Berber and S Asians looked like genetically at the time of Stuttgart. However, the problem is SSA and ENA are old signals, and are ancestral to other components. To neutralize the effect of SSA, I would have to replace some other component in S Asians and Stuttgart with it. The reverse would not make much sense. How do you suggest the effect be neutralized?

John Doe
04-26-2015, 04:34 PM
There are a couple of AJs close to you, but the majority seem more ME shifted. You may want to look at the k6 spreadsheet to compare with them. Not sure what region those AJs are from

Some are closer to me, but the bulk are in the gap between eastern Sicilians and Armenians. Personally my family comes from Germany and Poland, don't know if it means much though, as AJs from Alsace in the west to Minsk in the east tend to be as related as 5th cousins.

Just looked at the spreadsheet, on average my results are quite similar to those of other AJs, although I seem to be on the team of more Pre Yamnaya and less "Middle Eastern", but again, that's relative.

Dr_McNinja
04-26-2015, 08:55 PM
I presume what you mean is you would like to go back in time, and see what the Berber and S Asians looked like genetically at the time of Stuttgart. However, the problem is SSA and ENA are old signals, and are ancestral to other components. To neutralize the effect of SSA, I would have to replace some other component in S Asians and Stuttgart with it. The reverse would not make much sense. How do you suggest the effect be neutralized?

Perhaps just ignoring them altogether and recalculating percentages as if they weren't there? For example, 10% SSA, 10% ENA, 30% Yamnaya, 50% Pre-Yamnaya would become 37.5% Yamnaya and 62.5% Pre-Yamnaya.

But for people with heavy SSA/ENA, this would only represent a hypothetical portion of their admixture. Like an Oracle, as if they were a mix of two populations, the other being their ENA/SSA.

Kurd
04-26-2015, 11:01 PM
Perhaps just ignoring them altogether and recalculating percentages as if they weren't there? For example, 10% SSA, 10% ENA, 30% Yamnaya, 50% Pre-Yamnaya would become 37.5% Yamnaya and 62.5% Pre-Yamnaya.

But for people with heavy SSA/ENA, this would only represent a hypothetical portion of their admixture. Like an Oracle, as if they were a mix of two populations, the other being their ENA/SSA.

I can change the numbers any which way, but it would be very hypothetical, and not reflective of the situation in any time period.
A more realistic approach may be to go back in time to when the alleles associated with S Asian & E Eurasian had not formed yet, in other words, replace ENA with something ancestral to ENA. But this may take us further back than Stuttgart.

Arbogan
04-27-2015, 04:00 PM
How many snps is it based on? Anything below 100k snps can skew results.

J Man
04-27-2015, 04:59 PM
Here is the PCA. About 65% of the variation is captured by PC1/PC2. I have also posted the eigenvectors so that you have an idea of the component axes. The eigenvalues will give you an idea of the variation that is captured (last graphic)


http://i.imgur.com/L45aqS2.jpg
http://i.imgur.com/le3wr3M.jpg
http://i.imgur.com/Cby66tx.jpg

Excellent work Kurd it all looks quite good. I am a bit closer to Corded Ware and Yamnaya than I though I would be. Cool stuff.

Kurd
04-27-2015, 06:12 PM
How many snps is it based on? Anything below 100k snps can skew results.
I believe someone had mentioned 55k, but maybe someone can check to verify

Sein
04-28-2015, 05:09 AM
I have a few results to present.

Kandahari Afghan Pashtun (Durrani):

Middle_Eastern=54.7184%
Yamnaya_related=32.5316%
ENA=10.1279%
Sub-Saharan=1.4936%
WHG_Extra=1.1275%

Ghilji Afghan Pashtun:

Middle_Eastern=51.3947%
Yamnaya_related=36.2949%
ENA=10.4857%
Sub-Saharan=1.478%
WHG_extra=0.3456%

50% Laghmani Afghan Pashtun + 50% Panjsheri Tajik:

Middle_Eastern=48.8335%
Yamnaya_related=34.7916%
ENA=12.0205%
Sub-Saharan=2.5218%
WHG_extra=1.8315%

75% Afghan Pashtun (Nangarhar + Paktia) + 25% Pakistani Pashtun (Afridi)

Middle_Eastern=49.4086%
Yamnaya_related=36.0266%
ENA=12.0899%
Sub-Saharan=1.731%
WHG_extra=0.7429%

Interesting differences, although all are fairly similar.

The Kandahari Pashtun is the most western-shifted Pashtun I've seen yet, on calculators like HarappaWorld. Their "Caucasian" score is 25%, their "South_Indian" score is 15%, and their "Mediterranean" score is 3%. Despite this, their "ENA" score is scarcely distinguishable from other Pashtuns. I guess the Durrani of Kandahar are shifted further to the west primarily due to lower steppe input related to Indo-Europeans, and also higher recent West Asian gene-flow from further west on the Iranian plateau.

surbakhunWeesste
04-28-2015, 05:27 AM
I have a few results to present.

Kandahari Afghan Pashtun (Durrani):

Middle_Eastern=54.7184%
Yamnaya_related=32.5316%
ENA=10.1279%
Sub-Saharan=1.4936%
WHG_Extra=1.1275%

Interesting differences, although all are fairly similar.

The Kandahari Pashtun is the most western-shifted Pashtun I've seen yet, on calculators like HarappaWorld. Their "Caucasian" score is 25%, their "South_Indian" score is 15%, and their "Mediterranean" score is 3%. Despite this, their "ENA" score is scarcely distinguishable from other Pashtuns. I guess the Durrani of Kandahar are shifted further to the west primarily due to lower steppe input related to Indo-Europeans, and also higher recent West Asian gene-flow from further west on the Iranian plateau.

A quick question, where did you get the result from :)

Sein
04-28-2015, 05:36 AM
A quick question, where did you get the result from :)

I have their raw-data.

surbakhunWeesste
04-28-2015, 05:37 AM
I have their raw-data.

How did you get it? Did someone sent you my raw data? :O because that's my result.

Sein
04-28-2015, 05:39 AM
How did you get it? Did someone sent you my raw data? :O because that's my result.

That's impossible, this is a dude. He sent me his 23andMe raw-data quite awhile back, when I talked to him at Pashtunforums.

surbakhunWeesste
04-28-2015, 05:39 AM
That's impossible, this is a dude. He sent me his 23andMe raw-data quite awhile back, when I talked to him at Pashtunforums.

:O wut!


Yamnaya_related 0.325316
WHG_extra 0.011275
ENA 0.101279
Middle_Eastern 0.547184
Pre-Yamnaya 0.00001
Sub-Saharan 0.014936

Pop ID Oceanian European EEF East_Asian South_Asian Central_Asian Amerindian Siberian Sub-Saharan
0.006582 0.09646 0.110458 0.01904 0.20775 0.52418 0.003033 0.032487 1E-005

That's weird O_0

Sein
04-28-2015, 05:44 AM
:O wut!


Yamnaya_related 0.325316
WHG_extra 0.011275
ENA 0.101279
Middle_Eastern 0.547184
Pre-Yamnaya 0.00001
Sub-Saharan 0.014936

Pop ID Oceanian European EEF East_Asian South_Asian Central_Asian Amerindian Siberian Sub-Saharan
0.006582 0.09646 0.110458 0.01904 0.20775 0.52418 0.003033 0.032487 1E-005

That's weird O_0

This is amazing, your results are completely identical!

But I know this can't be you, since you have a separate 23andMe account from his.

Also, I only know he is from Kandahar, he hasn't told me that he is Durrani (but I'm 100% sure that he is Durrani).

He is an Admin at that forum.

paulgill
04-28-2015, 05:54 AM
I was also surprised. I am also still pondering ENA, and trying to account for S Asian admixture. Not sure whether Davidski's analysis or calculator may be a little off.


This is amazing, your results are completely identical!

But I know this can't be you, since you have a separate 23andMe account from his.

Also, I only know he is from Kandahar, he hasn't told me that he is Durrani (but I'm 100% sure that he is Durrani).

He is an Admin at that forum.Strange, may be he somehow got excess to her data, or the same data was assigned to two different kits.

Sein
04-28-2015, 05:59 AM
Strange, may be he somehow got excess to her data, or the same data was assigned to two different kits.

I told zahra his username, she is familiar with him. But he isn't a relative of her's, and she has never sent him her data.

I guess this can just be chalked up to chance.

paulgill
04-28-2015, 06:03 AM
I told zahra his username, she is familiar with him. But he isn't a relative of her's, and she has never sent him her data.

I guess this can just be chalked up to chance.Yes, it can be checked easily as her data should have no Ydna data, strange nevertheless.

Sein
04-28-2015, 06:12 AM
Yes, it can be checked easily as her data should have no Ydna data, strange nevertheless.

I just checked the file, it has Y-DNA data. Definitely a male individual's raw-data.

paulgill
04-28-2015, 06:33 AM
I just checked the file, it has Y-DNA data. Definitely a male individual's raw-data.I think Zahra should check her file too, just to make sure.

Generalissimo
04-28-2015, 07:05 AM
Copy paste errors do happen occasionally (actually, only once so far to be precise), but this appears to be the same file.

surbakhunWeesste
04-28-2015, 07:28 AM
Yeah! its all cool.

surbakhunWeesste
04-28-2015, 10:01 AM
I have a few results to present.

Kandahari Afghan Pashtun (Durrani):

Middle_Eastern=54.7184%
Yamnaya_related=32.5316%
ENA=10.1279%
Sub-Saharan=1.4936%
WHG_Extra=1.1275%


The Kandahari Pashtun is the most western-shifted Pashtun I've seen yet, on calculators like HarappaWorld. Their "Caucasian" score is 25%, their "South_Indian" score is 15%, and their "Mediterranean" score is 3%. Despite this, their "ENA" score is scarcely distinguishable from other Pashtuns. I guess the Durrani of Kandahar are shifted further to the west primarily due to lower steppe input related to Indo-Europeans, and also higher recent West Asian gene-flow from further west on the Iranian plateau.

Anyway, away from the fiasco and back to topic:

The Mediterranean score is actually .3% not 3%
Here is the result, Jesus posted it sometimes back with my consent.

http://www.anthrogenica.com/showthread.php?1229-Harappa-Ancestry-Project&p=65224&viewfull=1#post65224

We do not have any recent ancestry from Iran. My paternal side is Sadozai which as you know comes under Zirak. My maternal side is Alakozai so is my father's maternal side, whilst my mother's maternal side is Sadozai. This tandem goes back quite some generations. As you know we are a patriarchal society and esp. we hail from Kandahar that means if there is any Iranian ancestry it would be from the maternal side. Now, having said that my father's mtdna is M5a2a and mine is U2b2 which are almost completly absent if not absent at all in Iran. Most of our relatives show similar results as well.

The Barnacle
04-28-2015, 10:47 AM
I have a few results to present.

Kandahari Afghan Pashtun (Durrani):

Middle_Eastern=54.7184%
Yamnaya_related=32.5316%
ENA=10.1279%
Sub-Saharan=1.4936%
WHG_Extra=1.1275%

Ghilji Afghan Pashtun:

Middle_Eastern=51.3947%
Yamnaya_related=36.2949%
ENA=10.4857%
Sub-Saharan=1.478%
WHG_extra=0.3456%

50% Laghmani Afghan Pashtun + 50% Panjsheri Tajik:

Middle_Eastern=48.8335%
Yamnaya_related=34.7916%
ENA=12.0205%
Sub-Saharan=2.5218%
WHG_extra=1.8315%

75% Afghan Pashtun (Nangarhar + Paktia) + 25% Pakistani Pashtun (Afridi)

Middle_Eastern=49.4086%
Yamnaya_related=36.0266%
ENA=12.0899%
Sub-Saharan=1.731%
WHG_extra=0.7429%

Interesting differences, although all are fairly similar.

The Kandahari Pashtun is the most western-shifted Pashtun I've seen yet, on calculators like HarappaWorld. Their "Caucasian" score is 25%, their "South_Indian" score is 15%, and their "Mediterranean" score is 3%. Despite this, their "ENA" score is scarcely distinguishable from other Pashtuns. I guess the Durrani of Kandahar are shifted further to the west primarily due to lower steppe input related to Indo-Europeans, and also higher recent West Asian gene-flow from further west on the Iranian plateau.

Thanks Sein! Could you pm me the laghmani panjshiris Harappa results please? Or on any other calculator? I'm 25% laghmani :)

Btw have you got any western afghan results?
Tbh HRP0286 is 26% Caucasian and had 2.72% Med,

Where is the Ghilji pashtun from? Also, where would western afghans from Herat Farah nimruz etc stand?

The Barnacle
04-28-2015, 10:56 AM
Anyway, away from the fiasco and back to topic:

The Mediterranean score is actually .3% not 3%
Here is the result, Jesus posted it sometimes back with my consent.

http://www.anthrogenica.com/showthread.php?1229-Harappa-Ancestry-Project&p=65224&viewfull=1#post65224

We do not have any recent ancestry from Iran. My paternal side is Sadozai which as you know comes under Zirak. My maternal side is Alakozai so is my father's maternal side, whilst my mother's maternal side is Sadozai. This tandem goes back quite some generations. As you know we are a patriarchal society and esp. we hail from Kandahar that means if there is any Iranian ancestry it would be from the maternal side. Now, having said that my father's mtdna is M5a2a and mine is U2b2 which are almost completly absent if not absent at all in Iran. Most of our relatives show similar results as well.

If there is any Iranian ancestry among afghans then it would have to be the Qizilbash minority, and Persian speaking shias( not hazaras most likely Tajiks and Pashtuns, Khalid hosseini for example). I would love to send my DNA to find out if I have any recent ancestry from Iran because as you know my paternal grandmother was a Persian speaking Shia from Kandahar ( popalzai). I don't know the process though :(

The Barnacle
04-28-2015, 11:46 AM
Yeah! its all cool.

That can't be chance, the results are litetally identical. There must be some confusion LOL

Dr_McNinja
04-28-2015, 04:17 PM
I have a few results to present.

Kandahari Afghan Pashtun (Durrani):

Middle_Eastern=54.7184%
Yamnaya_related=32.5316%
ENA=10.1279%
Sub-Saharan=1.4936%
WHG_Extra=1.1275%

Ghilji Afghan Pashtun:

Middle_Eastern=51.3947%
Yamnaya_related=36.2949%
ENA=10.4857%
Sub-Saharan=1.478%
WHG_extra=0.3456%

50% Laghmani Afghan Pashtun + 50% Panjsheri Tajik:

Middle_Eastern=48.8335%
Yamnaya_related=34.7916%
ENA=12.0205%
Sub-Saharan=2.5218%
WHG_extra=1.8315%

75% Afghan Pashtun (Nangarhar + Paktia) + 25% Pakistani Pashtun (Afridi)

Middle_Eastern=49.4086%
Yamnaya_related=36.0266%
ENA=12.0899%
Sub-Saharan=1.731%
WHG_extra=0.7429%

Interesting differences, although all are fairly similar.

The Kandahari Pashtun is the most western-shifted Pashtun I've seen yet, on calculators like HarappaWorld. Their "Caucasian" score is 25%, their "South_Indian" score is 15%, and their "Mediterranean" score is 3%. Despite this, their "ENA" score is scarcely distinguishable from other Pashtuns. I guess the Durrani of Kandahar are shifted further to the west primarily due to lower steppe input related to Indo-Europeans, and also higher recent West Asian gene-flow from further west on the Iranian plateau.I dunno, I've inquired about HRP0370's ancestry and they're all Kandahar-area Durrani Pashtun on all sides (well, it becomes mostly one side with a couple of outsiders (non-Popalzai) marrying in once you go back a few generations because of inbreeding) going back to Ahmad Shah. Bamezai and Sadozai mostly with an Alakozai woman once or twice.

Her WHG is usually pretty high which made me guess some kind of Tajik ancestry (like Rukha who usually has high Euro/WHG) and inquire but there's no record of it. Plus on PCA plots they're where the other Afghan Pashtun are, pointing in the direction of the Caucasus/Iran whereas Rukha is very obviously shifted between Pashtun and Tajik.

Their J1b7 mtDNA could be Iranian though. It's only been found in Iraq/Turkey so far out of the J1b subclades.

http://www.eupedia.com/europe/Haplogroup_J_mtDNA.shtml#subclades

http://www.ianlogan.co.uk/sequences_by_group/j1b_genbank_sequences.htm (that is HRP0370's data I submitted to Ian)

If they're really above the norm for that part of Afghanistan, it could be the extra bit of WHG is from South Asia since even their ANI/Hindu-Kush is higher. Or just a fluke of inheritance and inbreeding (kind of like the archaic admixture in my family which now everyone has lots of, me above all).

Sein
04-28-2015, 05:18 PM
Anyway, away from the fiasco and back to topic:

The Mediterranean score is actually .3% not 3%
Here is the result, Jesus posted it sometimes back with my consent.

http://www.anthrogenica.com/showthread.php?1229-Harappa-Ancestry-Project&p=65224&viewfull=1#post65224

We do not have any recent ancestry from Iran. My paternal side is Sadozai which as you know comes under Zirak. My maternal side is Alakozai so is my father's maternal side, whilst my mother's maternal side is Sadozai. This tandem goes back quite some generations. As you know we are a patriarchal society and esp. we hail from Kandahar that means if there is any Iranian ancestry it would be from the maternal side. Now, having said that my father's mtdna is M5a2a and mine is U2b2 which are almost completly absent if not absent at all in Iran. Most of our relatives show similar results as well.

Yeah, I was going on memory (I ran this file through HarappaWorld quite a few months ago).

Also, I don't think your family is unique in having experienced some sort of gene-flow from further west on the Iranian plataeu. It's a process that must have effected most Durrani Pashtuns. That would explain why the Yamnaya-percentage is much lower than other Pashtuns, even though the ENA is pretty much the same.

The same applies to ANE, as HRP0370 had a very low ANE score for a Pashtun, using the calculator at GedMatch. So one needs gene-flow from ANE/steppe-ancestry poor groups that still have comparable ENA to Pashtuns. West Asians fit that bill, and the Durrani are the most geographically western (and southern) Pashtun tribal group.

jesus
04-28-2015, 07:05 PM
Yeah, I was going on memory (I ran this file through HarappaWorld quite a few months ago).

Also, I don't think your family is unique in having experienced some sort of gene-flow from further west on the Iranian plataeu. It's a process that must have effected most Durrani Pashtuns. That would explain why the Yamnaya-percentage is much lower than other Pashtuns, even though the ENA is pretty much the same.

The same applies to ANE, as HRP0370 had a very low ANE score for a Pashtun, using the calculator at GedMatch. So one needs gene-flow from ANE/steppe-ancestry poor groups that still have comparable ENA to Pashtuns. West Asians fit that bill, and the Durrani are the most geographically western (and southern) Pashtun tribal group.

Based on his K9 results, the west asian or north caucasian affinity seemed like it was lumped in the Central asian component (peaks in Caucasians) not into his EEF(which is mainly near eastern with tiny WHG). Her EEF is actually lower than the average Afghan Pashtun (10.7%), Afghan Tajik(12.7%) and Afghan Turkmen (14.7%). Zahra's father central asian (52.4%) is higher than Tajik Pamiri(45.3%), Afghan Tajik(38.5%), Afghan Pashtun(45%), Pathans(44.6%) and Iranians(46.7%). He scored the most central asian amongst All the Afghan, Tajik, Pamiri and Pakistani Pashtun samples.
West Iranian or Tajik ancestry would lower the Central Asian and increase the amount of EEF.
They seem to be more North caucasian shifted than Iranian shifted, in the K9 oracle he is closer to Lezgins, Chechens, Kumyuk, Adygei, North_Ossetian, and Balkar than to Iranians. Which is not surprising because those people have very high Central asian.
http://www.anthrogenica.com/showthread.php?4277-Upload-your-K9-results&p=78154&viewfull=1#post78154
http://www.anthrogenica.com/showthread.php?4277-Upload-your-K9-results&p=78835&viewfull=1#post78835

https://docs.google.com/spreadsheets/d/1M2drQQD7jvwwaTkpFTuBLQXllkUTVNOIfa9FReH7V5M/edit#gid=644681253 K9 spreadsheet

jesus
04-29-2015, 01:34 AM
Based on his K9 results, the west asian or north caucasian affinity seemed like it was lumped in the Central asian component (peaks in Caucasians) not into his EEF(which is mainly near eastern with tiny WHG). Her EEF is actually lower than the average Afghan Pashtun (10.7%), Afghan Tajik(12.7%) and Afghan Turkmen (14.7%). Zahra's father central asian (52.4%) is higher than Tajik Pamiri(45.3%), Afghan Tajik(38.5%), Afghan Pashtun(45%), Pathans(44.6%) and Iranians(46.7%). He scored the most central asian amongst All the Afghan, Tajik, Pamiri and Pakistani Pashtun samples.
West Iranian or Tajik ancestry would lower the Central Asian and increase the amount of EEF.
They seem to be more North caucasian shifted than Iranian shifted, in the K9 oracle he is closer to Lezgins, Chechens, Kumyuk, Adygei, North_Ossetian, and Balkar than to Iranians. Which is not surprising because those people have very high Central asian.
http://www.anthrogenica.com/showthread.php?4277-Upload-your-K9-results&p=78154&viewfull=1#post78154
http://www.anthrogenica.com/showthread.php?4277-Upload-your-K9-results&p=78835&viewfull=1#post78835

https://docs.google.com/spreadsheets/d/1M2drQQD7jvwwaTkpFTuBLQXllkUTVNOIfa9FReH7V5M/edit#gid=644681253 K9 spreadsheet

His EEF was 11%, my bad. It's still very similar to the average afghan Pashtun sample and lower than the afghan tajik and Turkmen.

One last thing, Zahra's father had 0% pre Yamnaya which is pretty typical for Pashtuns and Tajiks, other than few tajik/Pamiri samples who were 1-4% Pre Yamnaya which could signal either recent foreign ancestry(in the last 2500 years) or it could just mean a more ancient connection with the middle east/Europe. Iranians are 9.5% Pre Yamnaya on average, Kurds are 11.1%, Azeris are 12.2%, Greeks are 43.4% and Middle eastern Jews and North caucasians usually score over 15% Pre Yamnaya. Any recent admixture from those peoples should logically increase the pre-Yamnaya component in his case and in other cases.
Those populations had this pre Yamnaya component for at least 3000 years old, since it's a mix of WHG-UHG and ENF.

jesus
04-29-2015, 02:20 AM
I dunno, I've inquired about HRP0370's ancestry and they're all Kandahar-area Durrani Pashtun on all sides (well, it becomes mostly one side with a couple of outsiders (non-Popalzai) marrying in once you go back a few generations because of inbreeding) going back to Ahmad Shah. Bamezai and Sadozai mostly with an Alakozai woman once or twice.

Her WHG is usually pretty high which made me guess some kind of Tajik ancestry (like Rukha who usually has high Euro/WHG) and inquire but there's no record of it. Plus on PCA plots they're where the other Afghan Pashtun are, pointing in the direction of the Caucasus/Iran whereas Rukha is very obviously shifted between Pashtun and Tajik.

Their J1b7 mtDNA could be Iranian though. It's only been found in Iraq/Turkey so far out of the J1b subclades.

http://www.eupedia.com/europe/Haplogroup_J_mtDNA.shtml#subclades

http://www.ianlogan.co.uk/sequences_by_group/j1b_genbank_sequences.htm (that is HRP0370's data I submitted to Ian)

If they're really above the norm for that part of Afghanistan, it could be the extra bit of WHG is from South Asia since even their ANI/Hindu-Kush is higher. Or just a fluke of inheritance and inbreeding (kind of like the archaic admixture in my family which now everyone has lots of, me above all).

J1b7 is also found in one Pakistani Pashtun and one Persian from Central Iran.

http://www.eupedia.com/forum/threads/25613-Identifying-the-original-Indo-European-mtDNA-from-isolated-settlements?p=424605&viewfull=1#post424605

A big number of Tajiks probably have similar origins to Pashtuns(East Iranic), Afghan Tajiks have 0.4% to 6.1% WHG while Afghan Pashtuns are 1% to 4% WHG. It's a substantial difference imo, I seriously doubt that all of these Pashtuns samples have recent Unknown Tajik ancestry.

surbakhunWeesste
04-29-2015, 03:05 AM
J1b7 is also found in one Pakistani Pashtun and one Persian from Central Iran.

http://www.eupedia.com/forum/threads/25613-Identifying-the-original-Indo-European-mtDNA-from-isolated-settlements?p=424605&viewfull=1#post424605

A big number of Tajiks probably have similar origins to Pashtuns(East Iranic), Afghan Tajiks have 0.4% to 6.1% WHG while Afghan Pashtuns are 1% to 4% WHG. It's a substantial difference imo, I seriously doubt that all of these Pashtuns samples have recent Unknown Tajik ancestry.

This is true for most Southern Pashtuns, be it Ghilzais /Durranis/ Ghurghusht. The ethnic tension when it comes to marriage can lead to horrendous aftermath there. I can only imagine the antiquity.I remember having the Pashtun mixing with other ethnicities conv. in some other thread where two other Afghan members replied it to be more of a tolerable phenomena in northern and eastern provinces.
Like I said earlier in this forum, I come from a pretty close knit tribal families on both my paternal and maternal side. If anything, personally I kinda am making a sad statement but in the past the Pashtuns and Tajiks/Hazaras/Uzbeks/Qizilbash... didn't really get along when it comes to marriage and all. The tension is still felt there. A lot of Durranis didn't/don't get along with other Durranis too esp in the Southern areas.

Nevertheless, I am thoroughly enjoying the discussion.

Raskolnikov
04-29-2015, 04:16 AM
Yamnaya-related 34.2292%
WHG-extra 0.7412%
ENA 14.6179%
Middle_Eastern 48.56%
Pre-Yamnaya 0.001%
Sub-Saharan 1.8507%

Sein
04-29-2015, 04:34 AM
Based on his K9 results, the west asian or north caucasian affinity seemed like it was lumped in the Central asian component (peaks in Caucasians) not into his EEF(which is mainly near eastern with tiny WHG). Her EEF is actually lower than the average Afghan Pashtun (10.7%), Afghan Tajik(12.7%) and Afghan Turkmen (14.7%). Zahra's father central asian (52.4%) is higher than Tajik Pamiri(45.3%), Afghan Tajik(38.5%), Afghan Pashtun(45%), Pathans(44.6%) and Iranians(46.7%). He scored the most central asian amongst All the Afghan, Tajik, Pamiri and Pakistani Pashtun samples.
West Iranian or Tajik ancestry would lower the Central Asian and increase the amount of EEF.
They seem to be more North caucasian shifted than Iranian shifted, in the K9 oracle he is closer to Lezgins, Chechens, Kumyuk, Adygei, North_Ossetian, and Balkar than to Iranians. Which is not surprising because those people have very high Central asian.
http://www.anthrogenica.com/showthread.php?4277-Upload-your-K9-results&p=78154&viewfull=1#post78154
http://www.anthrogenica.com/showthread.php?4277-Upload-your-K9-results&p=78835&viewfull=1#post78835

https://docs.google.com/spreadsheets/d/1M2drQQD7jvwwaTkpFTuBLQXllkUTVNOIfa9FReH7V5M/edit#gid=644681253 K9 spreadsheet

But in the context of K6, any Central Asian ancestry would increase the Yamnaya_related score. The Pamiri Tajiks are around 40%-45%, and a few reach almost 50%! The other Tajik groups hover around 40%, and the Kalash are around 40%. And as can be seen from the results of Sikh Jatt individuals (I have Sapporo in mind), many people from the greater Punjab are around 40% for this component.

So the lower scores seen among Durrani Pashtuns have to be reflective of gene-flow from the west. We're not really talking about actual intermarriage with Iranians from that country. Rather, isolation-by-distance dynamics, which constitute a very powerful force over time.

Also, Durrani "Middle_Eastern" scores are much higher than any Tajik group, and higher than Kalash+eastern Pashtuns.

Sein
04-29-2015, 05:02 AM
A side note, but Zahra's father had the highest Iranian portion I've ever seen for a Pashtun (pretty much around 50% Iranian), using Douglas McDonald's program. Now that I know this raw-data doesn't belong to the person I thought it did, I'll pm Zahra the full results.

jesus
04-29-2015, 05:36 AM
But in the context of K6, any Central Asian ancestry would increase the Yamnaya_related score. The Pamiri Tajiks are around 40%-45%, and a few reach almost 50%! The other Tajik groups hover around 40%, and the Kalash are around 40%. And as can be seen from the results of Sikh Jatt individuals (I have Sapporo in mind), many people from the greater Punjab are around 40% for this component.

So the lower scores seen among Durrani Pashtuns have to be reflective of gene-flow from the west. Were not really talking about actual intermarriage with Iranians from that country. Rather, isolation-by-distance dynamics, which constitute a very powerful force over time.

Also, Durrani "Middle_Eastern" scores are much higher than any Tajik group, and higher than Kalash+eastern Pashtuns.

That middle eastern and neolithic is probably ancient, and it's most likely common in North Caucasians and Iranians as well. As for him being 50% Iranian in the oracle, he usually shows something like North caucasian + Sindhi or Kalash, way more often than Iranian + Sindhi( and sometimes Baloch or Kalash) or something like that. Also we shouldn't take the oracle too seriously because it can give so many different results at similar distances. For example I show Pathan, Kalash and Afghan a lot in my Oracles + Some kind of levantine population but In Gedmatch + 23andme + FTDNA I have literally no relatives from Afghanistan or Pakistan(Other than one person who had iranian ancestry) or the Levant. Which probably means that the relation is very ancient.

The lower middle eastern in Eastern pashtuns and Kalash could be because of the proximity of the Kandahar region to the Gedrosia AKA Balochistan. That region was probably affected by neolithic farmers more than the areas of Eastern Pasthuns and Kalash, Balochis and Brahuis score more Middle Eastern usually more than Kalash and Eastern Pashtuns.

There is a Kalash sample who is 50.4% Middle eastern and 36.1% Yamnaya, not that different from Zahra's father results. 2 of the Pakistani Pashtuns results were very similar as well:

HGDP00214
Middle Eastern 52.6%
Yamnaya 32.4%

HGDP00218
Middle eastern 51.9%
Yamnaya 34%

Also, all the Pakistani Pashtun samples lack the Pre-Yamnaya component which is quite interesting.

surbakhunWeesste
04-29-2015, 05:57 AM
A side note, but Zahra's father had the highest Iranian portion I've ever seen for a Pashtun (pretty much around 50% Iranian), using Douglas McDonald's program. Now that I know this raw-data doesn't belong to the person I thought it did, I'll pm Zahra the full results.

:) Dera Manana. I am much interested into checking out the McDonald's result.

Sein
04-29-2015, 05:57 AM
That middle eastern and neolithic is probably ancient, and it's most likely common in North Caucasians and Iranians as well. As for him being 50% Iranian in the oracle, he usually shows something like North caucasian + Sindhi or Kalash, way more often than Iranian + Sindhi( and sometimes Baloch or Kalash) or something like that. Also we shouldn't take the oracle too seriously because it can give so many different results at similar distances. For example I show Pathan, Kalash and Afghan a lot in my Oracles + Some kind of levantine population but In Gedmatch + 23andme + FTDNA I have literally no relatives from Afghanistan or Pakistan(Other than one person who had iranian ancestry) or the Levant. Which probably means that the relation is very ancient.

The lower middle eastern in Eastern pashtuns and Kalash could be because of the proximity of the Kandahar region to the Gedrosia AKA Balochistan. That region was probably affected by neolithic farmers more than the areas of Eastern Pasthuns and Kalash, Balochis and Brahuis score more Middle Eastern usually more than Kalash and Eastern Pashtuns.

There is a Kalash sample who is 50.4% Middle eastern and 36.1% Yamnaya, not that different from Zahra's father results. 2 of the Pakistani Pashtuns results were very similar as well:

HGDP00214
Middle Eastern 52.6%
Yamnaya 32.4%

HGDP00218
Middle eastern 51.9%
Yamnaya 34%

Also, all the Pakistani Pashtun samples lack the Pre-Yamnaya component which is quite interesting.

For what it's worth, McDonald's program looks at PCA dimensions. So it's quite different from the Oracles we see at Gedmatch.

And Zahra's father is at 32.5% Yamnaya_related, and 55% Middle Eastern, which is still quite significant in the context of local variation, when compared to the somewhat distinctive Kalash sample (it's a safe bet that this is the Kalasha sample that is among the least inbred).

HGDP00214 is a very unique sample. For HarappaWorld, it's 7% "SW-Asian", which is extremely high for a Pashtun (in fact, I have yet to see a Pashtun sample with that level of "SW-Asian"). The "South-Indian" score is only 12%, which is still the lowest of any Pashtun I've seen. The East Asian components add up to 8%, which is very high in comparison to other HGDP Pashtun samples. The other HGDP Pashtun samples are pretty much at around 1%-3% total East Asian.

In short, this sample probably has some non-Pashtun ancestry, which is expected, as Parachinar is quite an interesting place when it comes to mixture (the fact that it's the only Shia Pashtun city contributes to the distinctiveness).

If you look at HGDP00218, it's results are also quite different from other HGDP samples, but much less than HGDP00214 (I'm referring to HarappaWorld).

Sein
04-29-2015, 06:01 AM
:) Dera Manana. I am much interested into checking out the McDonald's result.

Sure thing. :)

It's only fair.

Also, if I can find them, I'll send the dendograms Everest did awhile back, they all included your father's data.

paulgill
04-29-2015, 06:17 AM
:) Dera Manana. I am much interested into checking out the McDonald's result.They be something like 95% Afghanistan + 5% Irish etc. if you have 23andMe results, it be essentially the same, so you don't really need Dr. McDonald's results at all.

jesus
04-29-2015, 06:20 AM
For what it's worth, McDonald's program looks at PCA dimensions. So it's quite different from the Oracles we see at Gedmatch.

And Zahra's father is at 32.5% Yamnaya_related, and 55% Middle Eastern, which is still quite significant in the context of local variation, when compared to the somewhat distinctive Kalash sample (it's a safe bet that this is the Kalasha sample that is among the least inbred).

HGDP00214 is a very unique sample. For HarappaWorld, it's 7% "SW-Asian", which is extremely high for a Pashtun (in fact, I have yet to see a Pashtun sample with that level of "SW-Asian"). The "South-Indian" score is only 12%, which is still the lowest of any Pashtun I've seen. The East Asian components add up to 8%, which is very high in comparison to other HGDP Pashtun samples. The other HGDP Pashtun samples are pretty much at around 1%-3% total East Asian.

In short, this sample probably has some non-Pashtun ancestry, which is expected, as Parachinar is quite an interesting place when it comes to mixture (the fact that it's the only Shia Pashtun city contributes to the distinctiveness).

If you look at HGDP00218, it's results are also quite different from other HGDP samples, but much less than HGDP00214 (I'm referring to HarappaWorld).

I agree, variation is bound to happen when it comes to large ethnicities such as Pashtuns and others. HGDP00218 has quite unique results indeed, His SW asian and South Indian are similar Baloch levels. Maybe he's from a different tribe, do HGDP Pathans belong to same tribe? Were they used in 23andme's Pathans sample ? People here say they are from the kurram valley while 23andme says they are from Peshawar.

In McDonald's program I'll probably show up as 40-60% levantine + 40-60% Pathan or Baloch. Or maybe as over 90% Iranian + some South Indian population.

Edit: 23andme claims that the Source of the Pathan sample is HGDP but also labels them as Peshawari.

Sein
04-29-2015, 06:30 AM
I agree, variation is bound to happen when it comes to large ethnicities such as Pashtuns and others. HGDP00218 has quite unique results indeed, His SW asian and South Indian are similar Baloch levels. Maybe he's from a different tribe, do HGDP Pathans belong to same tribe? Were they used in 23andme's Pathans sample ? People here say they are from the kurram valley while 23andme says they are from Peshawar.

In McDonald's program I'll probably show up as 40-60% levantine + 40-60% Pathan or Baloch. Or maybe as over 90% Iranian + some South Indian population.

Edit: 23andme claims that the Source of the Pathan sample is HGDP but also labels them as Peshawari.

Most sources put them in Kurram Agency. There is variation as to where, as some put them in lower Kurram, some put them in (or around) Parachinar. At least one sample is from Parachinar, no doubts whatsoever, HGDP00220. This sample is a big outlier, and most projects tend to exclude it from ADMIXTURE analyses. This sample is clearly 50% Hazara, one of the parents have to be from that ethnicity. The total East Asian score is an impressive 30%! Parachinar has a huge Hazara community, and the Shia affiliations of the Pashtun population would make intermarriage an easy and natural process.

If the HGDP Pashtuns are really from Kurram, they are going to be either Turi or Bangash (or both), with some Mangal and Chamkanni as further possibilities.

The Barnacle
04-29-2015, 07:49 AM
What happened to Shenandoah? ( HRP0286). His results are pretty interesting, does anyone have his raw data?

Ashina
04-29-2015, 01:17 PM
Me/dad

Yamnaya_related 0.219175 / 0.207105
WHG_extra 0.003601 / 0.011406
ENA 0.118948 / 0.109547
Middle_Eastern 0.478394 / 0.48816
Pre-Yamnaya 0.164792 / 0.179646
Sub-Saharan 0.01509 / 0.004136

Kurd
04-29-2015, 02:06 PM
Me/dad

Yamnaya_related 0.219175 / 0.207105
WHG_extra 0.003601 / 0.011406
ENA 0.118948 / 0.109547
Middle_Eastern 0.478394 / 0.48816
Pre-Yamnaya 0.164792 / 0.179646
Sub-Saharan 0.01509 / 0.004136

Welcome to the forum:welcome: Your results are the most E Eurasian shifted Turkish results I have seen yet. I also noticed that you had posted your K9 results. By looking at both, it appears that ENA in K6 is much more E Eurasian heavy than S Asian.

It seems like David did not account for all S Asian admixture in K6, as I can't really tell where it went, unless he lumped it into Middle Eastern, which frankly would seem a little odd. My 15.5% S Asian from K9 could not be fully accounted for by the 6% ENA in K6.

Edit: the above affects mostly individuals with substantial S Asian admixture

Edit2: Perhaps not as odd, if David is trying to create an admixture profile that predates the formation of S Asian admixture, but then again, I am not sure whether Yamnaya existed at that point in time.

Ashina
04-29-2015, 02:18 PM
Welcome to the forum:welcome: Your results are the most E Eurasian shifted Turkish results I have seen yet. I also noticed that you had posted your K9 results. By looking at both, it appears that ENA in K6 is much more E Eurasian heavy than S Asian.

It seems like David did not account for all S Asian admixture in K6, as I can't really tell where it went, unless he lumped into Middle Eastern, which frankly would seem a little odd.

Edit: the above affects mostly to individuals with higher S Asian admixture

So what do you say about the rest of my results? Except for the high East Asian? Average or no? Haven't seen many other Turkish K6/8/9 results so can't really compare.

Kurd
04-29-2015, 02:27 PM
So what do you say about the rest of my results? Except for the high East Asian? Average or no? Haven't seen many other Turkish K6/8/9 results so can't really compare.

E Eurasian was the noticeable one. I will let you know if I notice anything else.

ashwinb
04-29-2015, 03:30 PM
I was told by David this one doesnt do well for south asians, nevertheless:

K6

Yamnaya_related 0.355498
WHG_extra 1E-005
ENA 0.19516
Middle_Eastern 0.413738
Pre-Yamnaya 1E-005
Sub-Saharan 0.035584

ashwinb
04-29-2015, 03:46 PM
Does the "Yamnaya related" stand for "ANE" in the k8 calculator?And ENA as ASI and "Middleeastern" as "near eastern".
IS there a significant difference between "Middleeastern" and "PreYamnaya", how much of PreYamnaya is attributed to neolithic farmers?Is the rest of it all pre-neolithic?

thrax
04-29-2015, 05:34 PM
Have I posted my results? Here they are:

Yamnaya_related 0.230871
WHG_extra 0.012434
ENA 0.01081
Middle_Eastern 0.339996
Pre-Yamnaya 0.405878
Sub-Saharan 0.00001

icebreaker
04-29-2015, 05:59 PM
K6
Yamnaya_related 0.20047
WHG_extra 0.000241
ENA 0.071531
Middle_Eastern 0.578613
Pre-Yamnaya 0.136752
Sub-Saharan 0.012393

halildemirezen
04-29-2015, 09:25 PM
where is this Yamnaya test on gedmatch?

J Man
04-29-2015, 09:32 PM
where is this Yamnaya test on gedmatch?

It is not on GEDmatch.

halildemirezen
04-29-2015, 09:39 PM
It is not on GEDmatch.

Errr, how? :)

J Man
04-29-2015, 09:56 PM
Errr, how? :)

You need to contact David at Eurogenes yourself. It costs $20.00.

Kurd
04-29-2015, 10:15 PM
I was told by David this one doesnt do well for south asians, nevertheless:

K6

Yamnaya_related 0.355498
WHG_extra 1E-005
ENA 0.19516
Middle_Eastern 0.413738
Pre-Yamnaya 1E-005
Sub-Saharan 0.035584

It does not do well for anyone with substantial S Asian admixture as I explained in a previous post. The reason being is that I don't believe any of the K6 components fully accounts for all of the individual's S Asian admixture. ENA is more E Eurasian heavy than S Asian. I am not sure what your K9 S Asian was, but for me, all of my ~15.5% K9 S Asian admixture was not accounted for in my K6 ~6.1% ENA.

I had initially thought that maybe the K6 showed the admixture profile at a point in time that predates the formation of the S Asian component, where the S Asian is lumped into a component ancestral to S Asian. I don't believe that to be the case any more though

ashwinb
04-30-2015, 12:45 AM
It does not do well for anyone with substantial S Asian admixture as I explained in a previous post. The reason being is that I don't believe any of the K6 components fully accounts for all of the individual's S Asian admixture. ENA is more E Eurasian heavy than S Asian. I am not sure what your K9 S Asian was, but for me, all of my ~15.5% K9 S Asian admixture was not accounted for in my K6 ~6.1% ENA.

I had initially thought that maybe the K6 showed the admixture profile at a point in time that predates the formation of the S Asian component, where the S Asian is lumped into a component ancestral to S Asian. I don't believe that to be the case any more though

This is my K9:
63.31% ANI
24.82% ASI
0.00% Caucasian
0.00% Mesolithic
0.00% Neolithic
5.80% East-Asian
0.00% Subsaharian
0.00% Near-East
0.02% Pygmy
0.96% Amerindian
3.22% Papuan
0.90% Siberian
0.98% Altaic
Do you see any big discrepancies in results ok k9 and k6?

Kurd
04-30-2015, 02:50 AM
This is my K9:
63.31% ANI
24.82% ASI
0.00% Caucasian
0.00% Mesolithic
0.00% Neolithic
5.80% East-Asian
0.00% Subsaharian
0.00% Near-East
0.02% Pygmy
0.96% Amerindian
3.22% Papuan
0.90% Siberian
0.98% Altaic
Do you see any big discrepancies in results ok k9 and k6?

I'm not sure which calculator results these are, but I need to see your K9 Teal results

Anglecynn
05-03-2015, 01:21 AM
Here's mine:

Yamnaya_related 0.402852
WHG_extra 0.034237
ENA 0.00019
Middle_Eastern 0.062645
Pre-Yamnaya 0.493865
Sub-Saharan 0.006211
------------------
Rounded and ordered:

Pre-Yamnaya: 49.4%
Yamnaya_related: 40.3%
Middle_Eastern: 6.3%
WHG_Extra: 3.4%
Sub-Saharan: 0.6%
ENA: 0%

Longbowman put together some maps for this on another forum, judging by that i'm just slightly below average for pre-Yamnaya and Yamnaya-related (2% and 1% respectively), with the difference being made up by 2% extra Middle eastern compared to average and the rest presumably between the other more minor components of WHG_Extra and Sub-Saharan.

Krefter
05-03-2015, 03:25 AM
Here's mine:

Yamnaya_related 0.402852
WHG_extra 0.034237
ENA 0.00019
Middle_Eastern 0.062645
Pre-Yamnaya 0.493865
Sub-Saharan 0.006211
------------------
Rounded and ordered:

Pre-Yamnaya: 49.4%
Yamnaya_related: 40.3%
Middle_Eastern: 6.3%
WHG_Extra: 3.4%
Sub-Saharan: 0.6%
ENA: 0%

Longbowman put together some maps for this on another forum, judging by that i'm just slightly below average for pre-Yamnaya and Yamnaya-related (2% and 1% respectively), with the difference being made up by 2% extra Middle eastern compared to average and the rest presumably between the other more minor components of WHG_Extra and Sub-Saharan.

The Middle Eastern component isn't consistent for Euros. I've seen inconsistencies between family members. Your extra Middle_Eastern which makes you differnt probably doesn't mean much. There is a track of Middle Eastern in West Europe though, that wasn't there with Unetice, Urnfield, and Iron age Britons.

Shuffle
05-12-2015, 07:00 PM
South-Germany, with probably a little higher than average Middle-Eastern:


Yamnaya_related WHG_extra ENA Middle_Eastern Pre-Yamnaya Sub-Saharan
0.335431 0.022066 0.01667 0.093092 0.528977 0.003764