afghan pashtun dna to be posted soon

[DMXX]

one of my colleagues has been working on analyzing 189 samples of afghan pashtun dna, apparently he has found something interesting he is telling me and he will post results very soon, once they are posted, i will post them here

Not sure if you are referring to me, Karachi (we have interacted on DNA-Forums numerous times) but nevertheless;

Since the Lacaub et al. data was published almost one year ago, I set out to investigate the Y-DNA STR results of Afghan Pakhtuns from both Kandahar and Mazar-e-Sharif.

Unfortunately, my work has hit several snags along the way, none of which have been my own doing.
Due to restrictions by the Journal of Genetic Genealogy, I cannot "leak" any results, otherwise they will not accept what I have devoted months towards. Compounding this is the recent resignation of JoGG's lead editor. In effect, my paper (which is practically complete) has been sitting untouched on my Desktop for months.

I can safely disclose that the Pakhtuns from Afghanistan certainly appear more West Eurasian (e.g. Europe-Mid-East) along the paternal lines compared to the presumed Peshawari samples, which matches geographical trends (and may be largely presumed based on this criterion alone). Other than that, I'm afraid the information cap prevents me from saying much more.

[Karachi]

^ oh hi DMXX, what are you doing here? anyways i haven't been in touch with dnaforums.com for a very long time. I am actually following Eurogenes project because those pakistani pashtun samples and other pakistani samples are listed in the spreadsheets, as you can can see in the above posts. Eurogenes is dividing the samples nicely and we are getting a much clear picture of Pakistani population (dna)

This (west central asian) component is very interesting and is the main component in all pakistani populations listed, the south asian in pakistanis is actually second, which is surprising since pakistan usually is considered a south asian country. Though for the lower casts with in punjab/sindh I am pretty sure their main component is probably more south asian then west central asian, since they are much local to south asia

Btw those pashtun samples are from Kurram valley apparently not peshawar, but I am assuming afghan samples will be more influenced from west asian/eurasian because of their location and history.

[DMXX]

^ oh hi DMXX, what are you doing here?

Here to learn about Pakhtun culture and language mostly, but pleased to see you have been keeping the fine folks here updated on our genetics discoveries.

anyways i haven't been in touch with dnaforums.com for a very long time. I am actually following Eurogenes project because those pashtun samples and other pakistani samples are listed in the spreadsheets, as you can can see in the above posts. Eurogenes is dividing the samples nicely and we are getting a much clear picture of Pakistani population (dna)

DNA-Forums is down, possibly forever. Other regulars speculate the administrator (GvdM) has run into personal troubles or is critically ill.

I too have been following the Eurogenes and been paying special attention to results from our part of the world. They are interesting, to say the least, although I feel the North Atlantic component may be a West Asian-like signal with a two-ended spread in Eurasia (North Europe and South Asia), whilst the centroid remains (have you seen the Chechen values?).

I was surprised to see my South Baltic score (I am IR2, 7.9%) tends to be greater than the Pashtun and Tajik average in these recent runs. Few Iranians and Kurds score above 5%. Why that's the case, I'm still uncertain.

This (west central asian) component is very interesting and is the main component in all pakistani populations, the south asian in pakistanis is actually the second, which is surprising since pakistan is considered a south asian country

West Central Asian is likely a Neolithic component associated with a broader West Asian/Caucasian group, in my opinion. I am swayed by the following;

- Propensity of WC. Asian in the Iranian plateau also
- Central Asia's autochthonous Paleolithic and Mesolithic populations were hunter-gatherers who also partook in fishing. No evidence of subsistence farming or any other practices which are conducive to large population expansions connecting the swathe of land between Iran, Pakistan and Tajikistan together. The only parsimonious explanation is the movement of Neolithic farmers out of Northwest Iran through to the Caspian coastlines, eventually settling in various Central Asian oases.

I am not surprised to see Pakistanis as a whole are roughly 30% South Asian only. Rudimentary 23andMe sharing always gave me the impression that Pakistanis are more West Eurasian than their Indian counterparts, regardless of caste.

Btw those pashtun samples are from Kurram valley apparently not peshawar, but I am assuming afghan samples will be more influenced from west asian/eurasian because of their location and history.

I personally took Simranjits' word that they were from Peshawar years ago. Could you please show me the source stating they were from Kurram?

Indeed, the Afghani Pakhtun samples will likely be more West Asian/Eurasian and less South Asian. Geography is a fine predictor of autosomal affinities.

[DaMalakandGhar]

Here to learn about Pakhtun culture and language mostly, but pleased to see you have been keeping the fine folks here updated on our genetics discoveries.
...
''''
''''

DMXX, have they defined somewhere what do they mean by different genetic-geographical characterizations like West Central Asian, North Asian, etc.? Furthermore, I will opine that Kurram Valley samples might have been influenced by the adjoining South Asian populations towards the South-East.

One interesting feature of Iranian pop is Caucasus chunk as the single largest component.

[RevolutionThroughReason]

Not sure if you are referring to me, Karachi (we have interacted on DNA-Forums numerous times) but nevertheless;

Since the Lacaub et al. data was published almost one year ago, I set out to investigate the Y-DNA STR results of Afghan Pakhtuns from both Kandahar and Mazar-e-Sharif.

Unfortunately, my work has hit several snags along the way, none of which have been my own doing.
Due to restrictions by the Journal of Genetic Genealogy, I cannot "leak" any results, otherwise they will not accept what I have devoted months towards. Compounding this is the recent resignation of JoGG's lead editor. In effect, my paper (which is practically complete) has been sitting untouched on my Desktop for months.

I can safely disclose that the Pakhtuns from Afghanistan certainly appear more West Eurasian (e.g. Europe-Mid-East) along the paternal lines compared to the presumed Peshawari samples, which matches geographical trends (and may be largely presumed based on this criterion alone). Other than that, I'm afraid the information cap prevents me from saying much more.

Could this possibly imply that people of Pakhtunkhwah are by in large Pashtunized South Asians? This would debackle the claim that Pashtuns are a homogenize monolithic ethnic group that many Pashtun nationalists attempt to portray.

[Wahe Guru]

Here to learn about Pakhtun culture and language mostly, but pleased to see you have been keeping the fine folks here updated on our genetics discoveries.
...
I was surprised to see my South Baltic score (I am IR2, 7.9%) tends to be greater than the Pashtun and Tajik average in these recent runs. Few Iranians and Kurds score above 5%. Why that's the case, I'm still uncertain.
...
I am not surprised to see Pakistanis as a whole are roughly 30% South Asian only. Rudimentary 23andMe sharing always gave me the impression that Pakistanis are more West Eurasian than their Indian counterparts, regardless of caste.
...
Indeed, the Afghani Pakhtun samples will likely be more West Asian/Eurasian and less South Asian. Geography is a fine predictor of autosomal affinities.

"I was surprised to see my South Baltic score (I am IR2, 7.9%) tends to be greater than the Pashtun and Tajik average in these recent runs. Few Iranians and Kurds score above 5%. Why that's the case, I'm still uncertain."
- May I ask why were you surprised that your "South Baltic" score is greater tahn the Pashtun and Tajik? Where are you from then if I can ask?

"I am not surprised to see Pakistanis as a whole are roughly 30% South Asian only. Rudimentary 23andMe sharing always gave me the impression that Pakistanis are more West Eurasian than their Indian counterparts, regardless of caste." - what are you basing this statement on? Dna testing results? Appearance? Location? Guessing?
Your statement cannot be considered to be that correct unless it is proved by very controlled and accurate testing. I feel testers do not always have good knowledge of the people being tested. For example I think Sindhi Hindus and Sindhi Muslims are likely to show up as similar, and if you do the testing more finely in the same sub-groups of Sindhi Hindus and Muslims then the results are likely to be closer.

Looking at the spreadsheet for the Eurogenes project the "IN1 to IN20" group is 53% "South Asian" and therefore 47% others. But we dont know from where in India these IN individuals are?
The IN_S group is 57% South Asian and 43% others. I assume these are "South Indians"? But from where in south India? What other details do we have about them? I do believe it would be possible to make a study with 50 "South Indian samples" and show that they are all very similar to Irish - if thats what was wanted to be done.

[Wahe Guru]

Here to learn about Pakhtun culture and language mostly, but pleased to see you have been keeping the fine folks here updated on our genetics discoveries.



DNA-Forums is down, possibly forever. Other regulars speculate the administrator (GvdM) has run into personal troubles or is critically ill.

I too have been following the Eurogenes and been paying special attention to results from our part of the world. They are interesting, to say the least, although I feel the North Atlantic component may be a West Asian-like signal with a two-ended spread in Eurasia (North Europe and South Asia), whilst the centroid remains (have you seen the Chechen values?).

I was surprised to see my South Baltic score (I am IR2, 7.9%) tends to be greater than the Pashtun and Tajik average in these recent runs. Few Iranians and Kurds score above 5%. Why that's the case, I'm still uncertain.



West Central Asian is likely a Neolithic component associated with a broader West Asian/Caucasian group, in my opinion. I am swayed by the following;

- Propensity of WC. Asian in the Iranian plateau also
- Central Asia's autochthonous Paleolithic and Mesolithic populations were hunter-gatherers who also partook in fishing. No evidence of subsistence farming or any other practices which are conducive to large population expansions connecting the swathe of land between Iran, Pakistan and Tajikistan together. The only parsimonious explanation is the movement of Neolithic farmers out of Northwest Iran through to the Caspian coastlines, eventually settling in various Central Asian oases.

I am not surprised to see Pakistanis as a whole are roughly 30% South Asian only. Rudimentary 23andMe sharing always gave me the impression that Pakistanis are more West Eurasian than their Indian counterparts, regardless of caste.



I personally took Simranjits' word that they were from Peshawar years ago. Could you please show me the source stating they were from Kurram?

Indeed, the Afghani Pakhtun samples will likely be more West Asian/Eurasian and less South Asian. Geography is a fine predictor of autosomal affinities.

OK sorry, you said you are IR2 in the spreadsheet. "Iranian" I guess.

[Wahe Guru]

Could this possibly imply that people of Pakhtunkhwah are by in large Pashtunized South Asians? This would debackle the claim that Pashtuns are a homogenize monolithic ethnic group that many Pashtun nationalists attempt to portray.

Are you thinking that Pashtuns from other side of Afghanistan would be much different to those of KP? Me thinks the differences would be tiny, if any.

A Nigerian and Kenyan have greater genetic differences with each other, whereas in comparison a "pure" Swedish person is much closer to a "pure" Australian native Aborigine. Africans are more genetically diverse. Not said by me but by a world genetics expert.

[DMXX]

DMXX, have they defined somewhere what do they mean by different genetic-geographical characterizations like West Central Asian, North Asian, etc.? Furthermore, I will opine that Kurram Valley samples might have been influenced by the adjoining South Asian populations towards the South-East.

These component descriptors are always named after the region or population they reach a maximum in. For instance, the "Southwest Asian" component in both the Eurogenes and Dodecad projects is named so given it reaches a maximum in Bedouin, Saudi and other Arabian Peninsular populations.

These component names are not pre-determined and are given after the output data has been manually inspected. Beta runs in the Eurogenes project confirm this, as some spreadsheets contain the expected headers (Component1/1, Component2/2 etc.)

Thank you for contributing your thoughts on what the Kurram Valley's significance to the Pashtun results. Most autosomal studies reveal that geography plays a huge role in the genetic make-up of ethnic groups. With that being said, linguistic affiliations clearly have a genetic basis to them. The following chart from Yunusbayev et al.'s The Caucasus as an asymmetric semipermeable barrier to ancient human migrations neatly depicts both of the genetic trends I am describing;



Underscoring the first point, Iranic-speaking Tajiks and Pashtuns are closer to Altaic-speaking Uzbeks and Turkmens rather than Iranians. The converse applies with Anatolian Turks.

With that being said, both Iranians and Anatolian Turks are displaced further east of their presumed genetic location (e.g. Turks should theoretically sit between Cypriots and Armenians), implying a significant component pulls them towards Central Asia, which corroborates the dominant languages spoken in each group.

Could this possibly imply that people of Pakhtunkhwah are by in large Pashtunized South Asians? This would debackle the claim that Pashtuns are a homogenize monolithic ethnic group that many Pashtun nationalists attempt to portray.

Yes and no.

These results are specific to a single Pashtun population sampled. As Karachi has strived to emphasise on several genetic forums, one cannot presume genetic uniformity across such a broad territory, regardless of the practice of Pashtunwali and the accompanying endogamy.

One analogous example I can cite is that of West Iranic-speaking Gilakis and Mazandaranis along the southern Caspian coast in Iran. Despite the fact that both populations speak related languages and are neighbours, there are fairly significant differences in Y-DNA Chromosome variation between the two. Granted, genetic drift of some form has presumably taken place, but the point stands; neighbouring related populations will differ in one capacity or another, regardless of the variable considered.

Nationalist claims are usually based on romanticist rhetoric, special pleading and/or gross misinterpretations of empirical data. I contend the specific claim of a genetically homogeneous Pashto-speaking world would probably be shattered once future populations (Kandahar? Mazar-e-Sharif? Quetta? Chitral?) are tested autosomally. Pashtun history, particularly the aspect of Pashtunization, lends towards at least partial genetic diversity rather than absolute homogeneity.

With that being said, it may well be possible that Pashtuns regardless of geographic position are more similar with one another than their immediate neighbours. Indeed, this can be observed within Iranians regardless of their regional ancestry, or additionally, between Iranians and Kurds. Based on this, I am inclined to suspect the same will hold true of Pashtuns, particularly given their greater tendency to practice endogamy than their distant West Iranic cousins.

- May I ask why were you surprised that your "South Baltic" score is greater tahn the Pashtun and Tajik? Where are you from then if I can ask?

In several ADMIXTURE runs both Pashtuns and Tajiks have revealed themselves to have significantly greater North European components than Iranians.

You have correctly ascertained my background.

what are you basing this statement on? Dna testing results? Appearance? Location? Guessing?

As I stated it was from "rudimentary 23andMe testing", I based that statement on DNA testing results. Solely, I might add.

Your statement cannot be considered to be that correct unless it is proved by very controlled and accurate testing. I feel testers do not always have good knowledge of the people being tested. For example I think Sindhi Hindus and Sindhi Muslims are likely to show up as similar, and if you do the testing more finely in the same sub-groups of Sindhi Hindus and Muslims then the results are likely to be closer.

While I am inclined to agree with your thoughts, I generally approach this entire subject on a "wait and see" basis. I don't have any firm thoughts on Sindhi internal variation until fine-line data presents itself.

Looking at the spreadsheet for the Eurogenes project the "IN1 to IN20" group is 53% "South Asian" and therefore 47% others. But we dont know from where in India these IN individuals are?
The IN_S group is 57% South Asian and 43% others. I assume these are "South Indians"? But from where in south India? What other details do we have about them? I do believe it would be possible to make a study with 50 "South Indian samples" and show that they are all very similar to Irish - if thats what was wanted to be done.[/COLOR]

You are correct.

Those broad cluster labels are only useful in making (similarly) broad conclusions. As expected per geography, South Indians are less West Eurasian than their northern compatriots. I rarely see any claims of greater depth than that, which pleases me (shows even the lowest intellectual common denominator of genetic enthusiasts realise intricate theorising is simply inapplicable with such simple designations - That, or they're simply not posting it!).

Your last claim is highly improbable. While results can be bent in particular directions regardless of intent (e.g. Keyser et al.'s claim that South Siberian R1a1a remains connected with Indo-Iranians matched East Europeans best qualitatively and quantitatively despite the fact their database was heavily skewed towards Europe and Central Asian haplotypes are deficient), we are seeing plenty of overlap between recent autosomal studies and work carried out by independent bloggers.

It would take plenty of intentional bias-rigging to connect 50 random Karnatakans with 50 random Cork denizens.

[Admin Khan]

DMXX,
That was a very informative post. Just wanted to thank you.

[Karpo]

Factually speaking, Pashtuns are Bani Israel. This is verified by multiple accounts and stories from time immemorial, names, and also the simple fact that Pashto language is unique to the entire region, having little to no similarity with any other language around it. Ongoing and conclusive research is proving this to be the case.

[BarakzaiAbdali]

These component descriptors are always named after the region or population they reach a maximum in. For instance, the "Southwest Asian" component in both the Eurogenes and Dodecad projects is named so given it reaches a maximum in Bedouin, Saudi and other Arabian Peninsular populations.

These component names are not pre-determined and are given after the output data has been manually inspected. Beta runs in the Eurogenes project confirm this, as some spreadsheets contain the expected headers (Component1/1, Component2/2 etc.)

Thank you for contributing your thoughts on what the Kurram Valley's significance to the Pashtun results. Most autosomal studies reveal that geography plays a huge role in the genetic make-up of ethnic groups. With that being said, linguistic affiliations clearly have a genetic basis to them. The following chart from Yunusbayev et al.'s The Caucasus as an asymmetric semipermeable barrier to ancient human migrations neatly depicts both of the genetic trends I am describing;



Underscoring the first point, Iranic-speaking Tajiks and Pashtuns are closer to Altaic-speaking Uzbeks and Turkmens rather than Iranians. The converse applies with Anatolian Turks.

With that being said, both Iranians and Anatolian Turks are displaced further east of their presumed genetic location (e.g. Turks should theoretically sit between Cypriots and Armenians), implying a significant component pulls them towards Central Asia, which corroborates the dominant languages spoken in each group.



Yes and no.

These results are specific to a single Pashtun population sampled. As Karachi has strived to emphasise on several genetic forums, one cannot presume genetic uniformity across such a broad territory, regardless of the practice of Pashtunwali and the accompanying endogamy.

One analogous example I can cite is that of West Iranic-speaking Gilakis and Mazandaranis along the southern Caspian coast in Iran. Despite the fact that both populations speak related languages and are neighbours, there are fairly significant differences in Y-DNA Chromosome variation between the two. Granted, genetic drift of some form has presumably taken place, but the point stands; neighbouring related populations will differ in one capacity or another, regardless of the variable considered.

Nationalist claims are usually based on romanticist rhetoric, special pleading and/or gross misinterpretations of empirical data. I contend the specific claim of a genetically homogeneous Pashto-speaking world would probably be shattered once future populations (Kandahar? Mazar-e-Sharif? Quetta? Chitral?) are tested autosomally. Pashtun history, particularly the aspect of Pashtunization, lends towards at least partial genetic diversity rather than absolute homogeneity.

With that being said, it may well be possible that Pashtuns regardless of geographic position are more similar with one another than their immediate neighbours. Indeed, this can be observed within Iranians regardless of their regional ancestry, or additionally, between Iranians and Kurds. Based on this, I am inclined to suspect the same will hold true of Pashtuns, particularly given their greater tendency to practice endogamy than their distant West Iranic cousins.



In several ADMIXTURE runs both Pashtuns and Tajiks have revealed themselves to have significantly greater North European components than Iranians.

You have correctly ascertained my background.



As I stated it was from "rudimentary 23andMe testing", I based that statement on DNA testing results. Solely, I might add.



While I am inclined to agree with your thoughts, I generally approach this entire subject on a "wait and see" basis. I don't have any firm thoughts on Sindhi internal variation until fine-line data presents itself.



You are correct.

Those broad cluster labels are only useful in making (similarly) broad conclusions. As expected per geography, South Indians are less West Eurasian than their northern compatriots. I rarely see any claims of greater depth than that, which pleases me (shows even the lowest intellectual common denominator of genetic enthusiasts realise intricate theorising is simply inapplicable with such simple designations - That, or they're simply not posting it!).

Your last claim is highly improbable. While results can be bent in particular directions regardless of intent (e.g. Keyser et al.'s claim that South Siberian R1a1a remains connected with Indo-Iranians matched East Europeans best qualitatively and quantitatively despite the fact their database was heavily skewed towards Europe and Central Asian haplotypes are deficient), we are seeing plenty of overlap between recent autosomal studies and work carried out by independent bloggers.

It would take plenty of intentional bias-rigging to connect 50 random Karnatakans with 50 random Cork denizens.

Interesting thing is that Pashtun oral history and royal court historians discuss movement of outsider ethnic groups into Pashtun areas. On incursions into Delhi, some tribesmen from Afghanistan also took wives from amongst the conquered. Ahmad Shah for instance married a mughal princess. In other instances, outsiders were imported for business reasons. For instance, Britishers verified Hindus having been imported into FATA for business/banking by the tribes themselves which allowed for stabilization of dealing with money matters between clans.

There is a lot of data out there on tribal lineages of the major Pashtun tribes. To gain the "purest" sample, why not access hereditary tribal maliks or gain the Afghan government's support for such a project which would then provide a way to get at such samples? It could potentially provide a way to measure the accuracy of the oral history regarding relationships between tribes and mass migrations of tribes.

On the Pakistani side of the durand line, Britishers kept very detailed records of local tribal leaders as well.

For instance, the DNA of the swati royals and the Safi tribal leaders and the Khattak tribal leader might diminish "noise."

With all the work and money going into it, it doesnt seem like it would add to the costs significantly.

My father in law told me that in Waziristan, for instance, in the yargu khel clan area, the lands are exactly positioned in such a way so as to allow for a family tree to be constructed.

I guess what I am saying is that how the data is collected seems key to interpreting results. If I walk into Jalalabad and randomly select 100 people and ask for the DNA it may give a view of the genetic origin, on average, of the populace... but does it provide a historical view into ethnic roots?

Just in the last 270 years there have been multiple mass migration events related to war. Even now, say if the economy in Kabul overtakes the economy in Peshawar... a population shift can occur where Pakistan nationals try to move to where the money is better and then a DNA snap shot summarizes a population genetic profile. But that doesnt mean that the local tribes in the rural areas surrounding Kabul share that genetic ancestory.

The KP region and the Afghan basin both have had different conquerers themselves (e.g. the Sikhs reached a certain point and never entered Afghanistan). So I guess what I am saying is that if you were to find a Afridi tribal Lord and take his DNA versus the common man's DNA, which one is more reflective of the Pashtun ethnic group's baseline genetic lineage?

It seems like the only way to eliminate noise and confounders is to attempt and correlate sampling with history.

For instance, there are some Russian soldiers who converted to Islam and are now living in Kandahar province. It is unlikely that someone would have randomly gotten one of their children's samples... but if it were included in the analysis, attempting to broadly interpret those results to all of the province might lead us to chase European windmills.

[roadrunner]

I think there's some ambuguity about when the Durand Line came into existence and the province of Pakhtunkhwa!!

Pashtuns on both sides of the border won't show much "DNA" difference on average.

Also this North Asian and Southwest Asian component is a load of blog nonsense which noone has defined.

[fzpz]

not this crap again. people pretending to be enthusiasts about such drivel can't even define what a pashtun/sindhi/punjabi is much rather what the supposed nations' dna is. There can be no representative sample. Nations are social and political constructs created by the europeans. I'll take the Islamic approach to this: Who Cares!

[badmash_khel]

Underscoring the first point, Iranic-speaking Tajiks and Pashtuns are closer to Altaic-speaking Uzbeks and Turkmens rather than Iranians. The converse applies with Anatolian Turks.

One thing to note here is that so called Altaic speaking uzbeks and Turkmens are not purely mongoloid and they are mostly a mix of Iranic and Turkic. As you may know Iranic tribes (Scythians) lived in Central Asia before the arrival of Turkic tribes (uzbeks, turkoman etc). There are aproximately 11 million Tajiks in Uzbekistan. Due to government favouritism towards uzbeks and other reasons Tajiks of Uzbekistan have been forced to declare their ethnicity as Uzbek which is also their nationality. The farsi speaking Tajiks are the biggest part of population in Samarqand and Bukhara, the two big cities in Uzbekistan. So samples taken from Uzbekistan can be a bit misleading. The people of Uzbekistan and Turkmenistan might be Altaic speaking but their genetic make is certainly not purely mongoloid. Just like Azeris there are many turkified Tajiks among them.

Its clear from history and DNA testings that Iranians have significant semitic J contribution which makes sense since Iranian tribes arrived in the Iranian plateu as immigrants from further north and east. There were already established populations in places like Khuzestan when persians arrived on the scene.

Would it be fair to say that pashtuns and Tajiks as a whole are closer to original Aryans than modern Iranians? Iranians have less R1a than both Tajiks and Pashtuns.