As per the Harvard database, there are 454 male samples dated between 3000BCE and 0CE in all the *stan countries north of India and Russia combined. Of these, 166 are R1a but none is R1a-Y2 or L657+ which is supposed to have been formed around 2600BCE from Steppe R1a-Z94 > Y3.
In Narasimhan et al 2019 supplement too, there are 250 male samples analyzed between the dates 3000BCE - 0 CE, and none of them is R1a-Y3+. These include all the Sintashta, Andronovo, Saka and descended culture samples, all of these have significant steppe_mlba admixture.
This is the modern country-wise frequency of R-Y3+ lineages as per YFull (23 countries, 13250 samples). As you can see, only the Indian subcontinent sees R1a-Y3+ samples, with an average of ~15-20% of all males sampled in Pakistan, India, Nepal, Bangladesh and Sri Lanka. The only thing common between these and the steppe R1a is a common ancestor who lived around 2600BCE. But if his descendant Y3 was born in the steppe, we should have seen much more Y3 in the ancient samples and in the modern distribution in the countries north of India. So where are they?
I have provided a solution in my previous article which I think is most likely, but I'm willing to hear some other plausible ideas and perspectives. Let me know in the comments.
19 comments:
Have you accounted for population replacements?
"Have you accounted for population replacements?"
Expand please.
So I think you know what my theory is regarding PIE, Y-DNA etc. How far away do you think you are from my theory?
In simple terms I believe the main region of genetic expansion is from Gujarat region, then Indus River and then into all directions, east into India and North into Steppe.
Anatolian id from Armenian region, with ancestry similar to Hajji Firuz. Hajji Firuz may have inputs from SC asia, due to a wshg component, however we don't have 6000bce samples from SC asia, so cant confirm.
Similarly, steppe_en and khvalynsk seem to have excess wshg from SC asia.
So, PIE is either from Armenia, or from SC asia. If you want to connect it to Gujarat, you will need 6000bce samples from Gujarat.
Oh right.
But i dont think we need samples from Gujarat from 6000bc or from anywhere else.
Steppe groups have greater Ong affinity than Iran_N and Anatolian. Because Steppe definitely did not penetrate as far as the Andaman Islands, these increased Ong affinity shows ancient contacts between NW South Asians and Steppe.
(This may also be true of Papuans and others where South Asians are closr and hence steppe is closer)
Gujaratis are closest on average to most populations globally.
We dont ACTUALLY know who did or didnt speak PIE in ancient times and all we have are genetic data. The data suggests global population spread from the Gujarat and NW regions.
If populations came from this region than this original dispersal point would have had all the ancient eurasian y dna like e, g, j, i and R1etc.
E g j and i would have left earlier and hence we find them in more distant and diverged eurasian pops. R1 then dominates this region and then expands out hence we find it in more recent and genetically closer to indian pops. So iran_n does not have R1 becuase they left the disperal point before R1 became dominant there. They have had more time away hence their autosome is more diverged. Steppe Dna has had more recent ancestry from south asia hence it has the more recently dominant r1 lineages and is also closer to ong.
This explains why most of the eurasian y dna lineages can be found in south asia. R1 in south asia is likely going to be greater the closer you get genetically to Gujaratis or Rors and the ijg etc lineages should be found further away, the same as outside india where r1 is found closest to steppe and the others as you move away
"Steppe groups have greater Ong affinity than Iran_N and Anatolian. Because Steppe definitely did not penetrate as far as the Andaman Islands, these increased Ong affinity shows ancient contacts between NW South Asians and Steppe."
This could just be because of additional ANE ancestry in steppe. ANE itself has east eurasian Onge-like ancestry. You will have to make a stronger case.
Fstats are not the final answer for these questions, because there are multiple reasons for the affinities shown by Fstats. qpAdm models are required for proof.
I meant how the people of Central Asia are Turkic or Mongolian so we don’t detect y3 in Central Asia because the people carrying them there are extinct
"I meant how the people of Central Asia are Turkic or Mongolian so we don’t detect y3 in Central Asia because the people carrying them there are extinct"
Ok but we have loads of pre-turko-mongolic samples from the region, none have Y3 as well, but lots of Z2124.
People from Central Asia didnt really go extinct they just migrated to Europe. Europe has lots of post or non-Steppe Central Asian DNA like Scythians and many many related peoples.
All this DNA is likely unaccounted for when we model Euros with Steppe+WHG+Anatolian Farmer. There should be significant Scythian and related ancestry there especially in the case of Northern and Eastern Europeans.
"R1a-Y2 or L657+ which is supposed to have been formed around 2600BCE from Steppe R1a-Z94 > Y3."
On what basis do you claim that R1a Y3 formed around 26000 BCE? TMRCA data its just a number. you need Y3 ancient DNA to have a credible estimate of its time of origin.
"On what basis do you claim that R1a Y3 formed around 26000 BCE? TMRCA data its just a number. you need Y3 ancient DNA to have a credible estimate of its time of origin."
Based on an assumption of Years per mutation, and number of sequential downstream mutations in the modern population. Dates are from Yfull(https://www.yfull.com/tree/R-Y3/) which has the best Y tree currently and massive dataset.
And other papers also estimate similar age (poznik et al 2016 amongst others). All of these are likely to be underestimations of age for R1a subclades (Ding et al 2021).
Z2124 is a brother branch of Y3 under Z93.Yfull gives the formation date 2900-2200bce with 95%CI, same as for Y3. We found a Z2124 sample in I0432 2925-2491 calBCE Poltavka, corroborating Yfull. So Y3 date is highly likely to be correct as well.
Added Y hg distribution for england, ireland, spain, france, italy and greece to the dashboard.
Vasistha wrote,
"So, PIE is either from Armenia, or from SC asia."
Armenia is possible linguistically (Gamkrelidze and Ivanov), but it is difficult to see for a non specialist how it ccould be possible genetically. Aren't steppe_en wshg hypothetical ancestries not actually people?
Jaydeep has used the term SC Asia often based on memory. Which modern city Islamabad or Kabul would be considered SC Asia?
Mayuresh
sC Asia is Turkmenistan/Tajikistan/uzbekistan
Added Norway, Finland, Serbia as well. Total of 23000 samples in the dashboard now.
There is J2a in EBA and MBA anatolia. J2a is pretty common in Iran-Turan-Indus region. J2a in anatolia came via CHG/IranN affinity.
If J2a is a marker of Indo-hittite population then what was the language of all these J2a population residing in the east?
Then there are J2a lines among Greeks and nearby population which could have brought IE influence into the regions. CHG/IranN autosomal is again present there before steppe marker.
I could count 14 colors starting with the label R1a. Out of these 12 are present in South Asia and all 14 are present if Iran is added to the mix. Clearly IIr has the most diversity of R1. The argument that the region is more populous, and that is why it has more diversity is unconvincing. The mutilations take time to occur. Therefore, the diversity indicates antiquity.
Ukraine is not that far behind in terms of diversity; about 9 or may be 10, compared to 12 in Southern Asia. However only 2 of these 10 dominate the mix in Ukraine. And these are the two that are presemt in traces in IIr.
So imh layperson opinion, both OIT/AMT (AIT) would require only a few people going either way and changing the language of a very large number of peoples. Linguistically, an untenable scenario.
How then is R1a even relevant to PIE homeland debate?
I could count 14 colors starting with the label R1a. Out of these 12 are present in South Asia and all 14 are present if Iran is added to the mix. Clearly IIr has the most diversity of R1. The argument that the region is more populous, and that is why it has more diversity is unconvincing. The mutilations take time to occur. Therefore, the diversity indicates antiquity.
Ukraine is not that far behind with diversity; about 9 or may be 10 compared to 12 in Southern Asia. However only 2 of them dominate the mix in Ukraine. These are exactly the two that are presemt in traces in IIr.
So imh layperson, both AIT/OIT would require only a few people going either way and changing the language of very large numbers of people. Linguistically, an untenable scenario.
How relevant is R1a to the PIE homeland debat then?
Saharia tribe has highest diversity for R1 subclades and population of saharia is half million,and would be even less in the past.
That population point is cope from them
Post a Comment