search this blog
Thursday, May 28, 2020
An early Mitanni?
I've updated my Global25 datasheets with most of the ancients from the new Skourtanioti et al. paper. Here's a Principal Component Analysis (PCA) based on the data. It was produced with the Vahaduo PCA tools freely available here and the text file here.
Note that one of the Bronze Age females from Alalakh, labeled ALA019, appears to have ancestry from Turan and the Eurasian steppe. She may well have been a Mitanni of Indo-Aryan origin.
Interestingly, a Copper Age male from Arslantepe, ART038, belongs to Y-haplogroup R1b1a2 aka R1b-V1636. This is an unusual find, because R1b hasn't yet been reported in any Copper Age or earlier samples from outside of Europe and the Eurasian steppe.
As far as I can tell, this individual doesn't harbor any genome-wide ancestry from north of the Caucasus. However, R1b-V1636 is a rare lineage that is first attested in Eneolithic samples from the North Caucasus Piedmont steppe, so ART038's Y-chromosome might be the first evidence of the presence of steppe ancestry in Copper Age Anatolia.
I've also added most of the ancients from the new Agranat-Tamir et al. paper to the Gobal25 datasheets. The PCA below is based on the text file available here.
The Megiddo samples include a trio of interesting outliers dated to 1600-1500 BCE with significant ancestry from the steppe. One of these individuals is a male, I2189, who belongs to Y-haplogroup R and probably R1a. So he might also be of Indo-Aryan origin.
Another Megiddo male, S10768, belongs to R1b-M269 and probably shows a few per cent of steppe ancestry. I've already discussed how R1b and steppe ancestry may have ended up in the Bronze Age Near East in a couple of my previous posts:
R1b-M269 in the Bronze Age Levant
How did steppe ancestry spread into the Biblical-era Levant?
R-V1636: Eneolithic steppe > Kura-Araxes?
Labels:
Anatolia,
ancient DNA,
Canaanite,
Caucasus,
Eastern Europe,
Indo-Aryan,
Indo-European,
Indo-Iranian,
Levant,
Mesopotamia,
Mitanni,
Pontic-Caspian steppe,
Proto-Indo-European,
R1a,
R1a-Z93,
R1b-M269,
R1b-V1636
Subscribe to:
Post Comments (Atom)
333 comments:
«Oldest ‹Older 201 – 333 of 333@Slumbery,
Rob is right about Krepost. Krepost is from a different phenomena than AnatoliaChalcolithic.
"BTW, funnily enough, if I ask Genoplot what are the samples in the database closest to Barcin Calcolithic, the top of the list is dominated by Rome_Imperial."
Yeah, I guess roughly they have similar ratios of ancient ancestry. It is weird how that works.
@ Slumberry
“ However Barcin_C itself is as old as early KAC and it is actually older than the oldest dated KAC samples we have in the G25 database. And in the westernmost part of Anatolia. ”
Correct. Which is why CrM & I were suggesting to something else- Sioni; or related to Majkop-Leila Tepe horizon.
Dovetails with the appearance of “Caucasus-style” arsenical Copper in western anatolia at this time
@Rob
OK, I was under the impression that someone in the discussion argued that the root of the CHG spread in Anatolia is KAC itself. Reading back again I can see that I misunderstood something. I also agree that Krepost_N is an odd-out sample and a theory built on that would be shaky. (It is still interesting what process brought it there though.)
@gamerzj, yes, the Krepost sample is interesting, particularly because it's at the very earliest neolithic in Europe (and less than a millennium after the Barcin samples average). Though we don't know what it really says about anything.
Some plots using Davidski's West Eurasia PCA trends over time: https://imgur.com/a/2PWgDZs
It does seem like after Krepost you have very few outliers that are shifted towards CHG specifically, but there is this one Vinca culture outlier that seems to me to be pretty similar to the TepecikCiftlik Neolithic (sample ref I1894 at 5150BCE per Harvard's lists of mean dates). Though this is Hungary.
Then at Bulgaria and Greece at 4000 BCE there could be a slight shift towards Anatolia of that era, after a shift of slightly earlier Bulgaria and Serbia samples towards HG, but it's hard to pick apart from HG introgression just being more limited.
The plots overall show HG shift, but this is probably exacerbated by the shift to sampling being later on in Western, Central and Northern Europe - compare plots in Rivollat's paper - https://advances.sciencemag.org/content/6/22/eaaz5344
It'll be good to add some more samples from Turkey over time to this sort of plot, and then work out more about whether the Southern Balkans shows more of a pattern of just limited HG introgression, or more of a pattern where there is a little bit more than thought, but also some limited ongoing exchange with Western Turkey that dampens this. I have no idea if archaeology says yes or no to either scenario.
Another quick punt at trying to shed light on the Megiddo outliers question.
Simulate a cline extension of the Megiddo 3 outliers, using regression method to account for noise, on the assumption that the Megiddo_MLBA end is fixed.
Pastebin of cline extensions: https://pastebin.com/F5dkTiUh (a couple of points on the same cline)
The point A (on the assumption that I10100 has 50% Megiddo_MLBA ancestry), is closest to Kubano-Tersk and Yamnaya-Caucasus.
The intermediate point on the cline (on assumption I10100 has 75% Megiddo_MLBA ancestry), B is closest in order to (of a small set that I compared to) top 3: IRN_Hajji_Firuz_BA:I4243, ARM_LBA:RISE407, TKM_IA:DA382. Some of that may just be that TKM_IA is slightly "north" of Haji_Firuz_BA, but
Plotting in PCA in Past3 shows that the cline extensions immediately appear equally to point towards Kubano-Tersk / Yamnaya_Caucasus and Zevakinsky_MLBA... but then are distinct as Zevakinsky_MLBA gets its eastern shift from excess of East Eurasian ancestry which the cline extension lacks.
Haji_Firuz_BA seems kind of early for this kind of ancestry (2326 BCE), but maybe its ancestral population persisted for a while. If the samples continued the trend of early male samples from the Near East and Western Iran with steppe ancestry having R1b-M269 and no sign of R1a (I think?), that would also be consistent.
When Vahaduo fitting using TKM_IA, Haji_Firuz_BA and Megiddo main average, Haji_Firuz_BA is preferred in two out three, but they are a brother and sister, and TKM_IA in the "most" steppe shifted I10100. So you could take that either way: https://imgur.com/a/kJ76Pzf
Proximal fits using those two don't seem much improved compared to fits using "steppe cline" sources directly: https://imgur.com/a/vSyd3Nt . Best distances are about the same.
Because the Megiddo set itself covers probably a range of interal variation in Barcin&LevantN related ancestry (and this is the main diff between Steppe_MLBA and Steppe_EMBA), it may be difficult to extend out from that.
It seems like a consensus is being reached that I2189 belongs to R1a.
https://anthrogenica.com/showthread.php?20461-The-Genomic-History-of-the-Bronze-Age-Southern-Levant&p=671583&viewfull=1#post671583
@Romulus,
I didn't know that about WHG. Some Scyhtians used to make scull-cups from human heads and I think Botai also practiced something similar.
"Herodotus first wrote of andropophagi in his Histories, where he described them as one of several tribes near Scythia. An extra note indicates that the andropophagi are cannibals, as reflected in their name:
The manners of the Androphagi are more savage than those of any other race. They neither observe justice, nor are governed, by any laws. They are nomads, and their dress is Scythian; but the language which they speak is peculiar to themselves. Unlike any other nation in these parts, they are cannibals.
— Histories, Book 4 (Melpomene), trans. George Rawlinson, 1858-1860
Pliny the Elder later wrote in his Naturalis Historia that the same cannibals near Scythia wore the scalps of men on their chest.
The Androphagi, whom we have previously mentioned as dwelling ten days' journey beyond the Borysthenes, according to the account of Isigonus of Nicæa, were in the habit of drinking out of human skulls, and placing the scalps, with the hair attached, upon their breasts, like so many napkins."
@All
I've updated the blog post to reflect the new findings.
I'll probably do more detailed posts about some of these samples in the near future.
@Nezih Seven
Thanks for the answer. Makes sense regarding the Kurds. I assume there was power vacuum the allowed them to move into Eastern Turkey from the Middle Zagros?
I do wonder what happened to the there Persians and Medians that settled in Eastern Turkey though. Absorbed by Armenians?
@Davidski
“It seems like a consensus is being reached that I2189 belongs to R1a”
Maybe some early Indo-Iranians:
https://i.postimg.cc/FFLWsDv2/Megiddo.jpg
@East Pole @Davidski It would be really exciting to find out why the Corded Ware Culture was overwhelmingly R1a while most other Indo-Europeans were R1b.
About the Megiddo outliers.
Interesting that the same models that work well for I2200 and reasonably well for I10100 are producing a really bad for for I2189. In G25 nMontes:
"sample": "Levant_Megiddo_MLBA:Average",
"distance": 1.2117,
"Alalakh_MLBA": 53,
"Levant_PPNB": 26.5,
"Hajji_Firuz_C": 20.5,
"Catacomb": 0,
"Sintashta_MLBA": 0,
"Areni_C": 0
"sample": "Levant_Megiddo_MLBA_o1:I2200",
"distance": 1.1816,
"Alalakh_MLBA": 28.5,
"Levant_PPNB": 5,
"Hajji_Firuz_C": 39,
"Catacomb": 9,
"Sintashta_MLBA": 17,
"Areni_C": 1.5
"sample": "Levant_Megiddo_MLBA_o2:I10100",
"distance": 2.8833,
"Alalakh_MLBA": 27.5,
"Levant_PPNB": 0,
"Hajji_Firuz_C": 31,
"Catacomb": 28,
"Sintashta_MLBA": 13.5,
"Areni_C": 0
"sample": "Levant_Megiddo_MLBA_o1:I2189",
"distance": 4.192,
"Alalakh_MLBA": 44,
"Levant_PPNB": 0.5,
"Hajji_Firuz_C": 20.5,
"Catacomb": 13.5,
"Sintashta_MLBA": 15,
"Areni_C": 6.5
@Davidski
“It seems like a consensus is being reached that I2189 belongs to R1a”
Maybe some early Indo-Iranians:
https://i.postimg.cc/FFLWsDv2/Megiddo.jpg
And maybe they went through Iran:
https://i.postimg.cc/WzJBwvQV/Megiddo2.jpg
@Slumbery, as we move through worse-to-better models: https://imgur.com/a/wBN0Vqy , we see that I2189 retains a higher distance. (These aggregated models that can pick which samples from the source populations best fit, not something I like to do commonly when looking at deep ancestry, but here we're looking for what is plausibly proximal.).
It's possible there is just some odd placement of the sample in some dimension we don't care about, for purely random error reasons (as his sister doesn't show it).
@Romulus
Nice info but it only talks about it in the Magdalenian zone, and Magdalenians are differentiated from WHG proper due to the Aurignacian component. Were Italian Epigravettians, and more specifically, Italian (central + southern) sites associated with this as well? Because those would be the zones where the purest WHGs have been found, not western and central Europe under Magdalenians.
One thing to add about the Megiddo trio, should we go with the model that posits Steppe_MLBA+IranCA like ancestry (it's not particularly strong compared to the Steppe_EMBA model but also not especially weak, and we've got an R1a after all), there isn't anything that necessarily means we would have to have any admixture taking place in the Bactria-Margiana sites.
IranCA ancestry would be present in the whole of Northern Iran, and it's fairly likely that an Mitanni-Aryan group would have admixed in Iran, en route before arriving in Levant. I would be cautious about the idea that TKM_IA like people existed in number in Central Asia, or anything like this. There may have been but we can't conclude that when there are alternative ways that this could happen that don't involve ancestry actually from the BMAC or IranN related groups in Central Asia.
@Davidski
It would be really interesting to see how genetically speaking the Iron age II possible Israelite from Upper Galilee, Abel Beth Maacah looks like. Would you be able to run G25 for him?
https://en.wikipedia.org/wiki/Battle_of_Megiddo_(15th_century_BC)
Some observations.
The Haji Firuz BA with a lot of steppe can't be really be an early Indo-Aryan, at circa 2300 BC it precedes even Sintashta. Some Gutians-like lost people seems the most plausible explanation IMO.
I don't understand the association of ALA019 with the Mitanni dynasty, on the other side of Mesopotamia you would have had many people like her, not too surprising to see someone like her there, and billions of reason that have nothing to do with the Mitanni.
The Megiddo trio is puzzling, their DNA doesn't scream BMAC to me (but they don't work well G25 for some reason), and they are similar to Haji Firuz BA that is almost 1000 years older. They are not great Mitanni candidates IMO but ultimately that association of Indo-Iranian with the Mitanni isn't the most clear cut thing in the universe so..
There's no particular proximity to IRN_Hajji_Firuz_BA, it doesn't hit any models at all. IRN_Hajji_Firuz_C is included in all models, it is related to the general ethnocultural background.
Distance to: Levant_Megiddo_MLBA_o1:I2189
0.03940140 IRN_Hajji_Firuz_C:I4351
0.04318738 IRN_Hajji_Firuz_C:I2323
0.04821079 BGR_EBA:I2165
0.04990551 Levant_Yehud_IBA:I7003
0.05010240 IRN_Hajji_Firuz_C:I4349
0.05351327 IRN_Hajji_Firuz_BA:I4243
0.05351439 Yamnaya_BGR:Bul4
0.05676742 IRN_Hajji_Firuz_C:I4241
0.06064380 BGR_IA:I5769
0.06318750 KAZ_Zevakinskiy_LBA:I4267
0.06584763 BGR_EBA:I2176
0.06618693 KAZ_Zevakinskiy_BA:I3770
0.06726879 KAZ_Zevakinskiy_LBA:I3772
0.06734872 BGR_Beli_Breyag_EBA:Bul6
0.06887518 KAZ_Zevakinskiy_LBA:I3976
0.06932741 BGR_Krepost_N:I0679_d
0.07094716 KAZ_Zevakinskiy_LBA:I4295
0.07162479 BGR_MP_N:I1108
0.07200507 KAZ_Zevakinskiy_MLBA:I3763
0.07297876 BGR_EBA:I2520
0.07300199 KAZ_Zevakinskiy_LBA:I3977
0.07415423 BGR_EBA:I2175
0.07538455 KAZ_Zevakinskiy_LBA:I3753
0.07582638 BGR_MLBA:I2163
0.07714817 BGR_Varna_En2:ANI159-ANI181
Distance to: Levant_Megiddo_MLBA:I2190
0.03989987 IRN_Hajji_Firuz_C:I2323
0.04190537 IRN_Hajji_Firuz_C:I4351
0.04504653 Levant_Yehud_IBA:I7003
0.04626327 IRN_Hajji_Firuz_C:I4241
0.04678183 IRN_Hajji_Firuz_C:I4349
0.05627033 BGR_EBA:I2165
0.05847085 BGR_IA:I5769
0.06034633 BGR_Beli_Breyag_EBA:Bul6
0.06228973 Levant_PPNB:I1707
0.06286804 Levant_PPNB:BAJ001
0.06409633 BGR_EBA:I2176
0.06578890 BGR_Krepost_N:I0679_d
0.06621699 Levant_PPNB:I1704
0.06712928 Levant_PPNC:I1699
0.06744605 BGR_EBA:I2520
0.06878634 IRN_Hajji_Firuz_BA:I4243
0.06881395 BGR_EBA:I2175
0.06950065 Yamnaya_BGR:Bul4
0.06984039 BGR_MP_N:I1108
0.07082281 Levant_PPNB:I1710
0.07164998 Levant_PPNB:I0867
0.07402722 BGR_Middle_C:I2431
0.07407908 BGR_MP_N:I1295
0.07414978 BGR_Varna_En2:ANI159-ANI181
0.07470977 BGR_Late_C:I2430
Distance to: Levant_Megiddo_MLBA:I2195
0.03309985 Levant_Yehud_IBA:I7003
0.03557345 IRN_Hajji_Firuz_C:I4351
0.03587409 IRN_Hajji_Firuz_C:I2323
0.04314441 IRN_Hajji_Firuz_C:I4241
0.04744270 BGR_IA:I5769
0.04764861 IRN_Hajji_Firuz_C:I4349
0.05215093 Levant_PPNB:I1707
0.05269658 BGR_Krepost_N:I0679_d
0.05539242 BGR_EBA:I2165
0.05640789 BGR_Beli_Breyag_EBA:Bul6
0.05656333 BGR_EBA:I2176
0.05718654 Levant_PPNB:I1704
0.05956215 Levant_PPNC:I1699
0.06109607 BGR_C:I2519
0.06184796 BGR_EBA:I2175
0.06266331 BGR_Late_C:I2427
0.06268923 BGR_EBA:I2520
0.06319383 Levant_PPNB:BAJ001
0.06352433 BGR_Varna_En2:ANI159-ANI181
0.06370934 BGR_Middle_C:I2431
0.06404795 BGR_Late_C:I2430
0.06505306 BGR_Late_C:I2424
0.06515988 Levant_PPNB:I0867
0.06598970 BGR_Late_C:I2425
0.06599530 BGR_N:I2529
Distance to: Levant_Megiddo_MLBA_o2:I10100
0.03786040 IRN_Hajji_Firuz_BA:I4243
0.04249953 IRN_Hajji_Firuz_C:I4349
0.04463418 Yamnaya_BGR:Bul4
0.04614466 IRN_Hajji_Firuz_C:I4351
0.04679498 IRN_Hajji_Firuz_C:I2323
0.04687963 BGR_EBA:I2165
0.04757804 KAZ_Zevakinskiy_LBA:I3976
0.04921900 KAZ_Zevakinskiy_LBA:I3977
0.04932656 KAZ_Zevakinskiy_BA:I3770
0.04943015 KAZ_Zevakinskiy_LBA:I4267
0.04943238 IRN_Hajji_Firuz_C:I4241
0.04954372 KAZ_Zevakinskiy_LBA:I4295
0.04974384 KAZ_Zevakinskiy_MLBA:I3763
0.05005387 KAZ_Zevakinskiy_LBA:I3772
0.05133439 BGR_MLBA:I2163
0.05503354 KAZ_Zevakinskiy_LBA:I3753
0.05742778 BGR_IA:I5769
0.06463869 RUS_Afanasievo:I3388
0.06511313 RUS_Afanasievo:I5273
0.06629495 Levant_Yehud_IBA:I7003
0.06733097 RUS_Afanasievo:I5269
0.06792893 RUS_Afanasievo:I11112
0.06804028 RUS_Afanasievo:I10564
0.06827357 RUS_Afanasievo:I2071
0.06827496 RUS_Afanasievo:I6715
@Rob,
"Just think about this more- I still think Majkop & KAC are distinctive phenomena. KAC seems to be a heterogeneous network of traders, farmers, pastoralists.? I think with more y-dna sampling; the difference between them will be more apparent
The shift in MCA Anatolia goes back to Sioni horizon- that’s what Ikiztepe shows- which is even before KAC. Some KAC people might have taken Arslantepe itself; however"
The main Kura-Araxes homeland Y-DNA lineages are likely to be certain clades of J2a-M67, J1-Z1842 and R1b-V1636...Of course others may show up too though in smaller amounts. Hopefully more samples will be tested at some point.
@Slumberry
Between 25- 40% steppe ancestry megido samples that huge !!!!
And im starting to understand where all the indo iranian influence in proto hebrews came from
Have you ever noticed that Abraham and his wife Sarah of the old testament are nearly identical in name to Brahma and his wife Saraswati of the Hindu religion?
Abraham is said to be the father of the Jews, and Brahma, as the first created being, is often seen as the father of mankind…’ We might also note that the name of Brahma’s consort Sarasvati seems to resonate with that of Abraham’s wife, Sarah [… each one’s identity as a wife and/or sister]. Also, in India, the Sarasvati River includes a tributary known as the Ghaggar…. According to Jewish tradition, Hagar was Sarah’s maidservant
Abraham/Sarah and Brahma/Saraswati are not the only overlapping figures between Hindu and Hebrew traditions. Ajit Vadakayil points out that there is also Adam/Eve and Adhama/ Havyavati, along with Noah and Nyuha or Manu
Brahma as Father of All , while Abraham as Father of many nations
Brahman Milky Way is the Celestial Cow/Dolphin; with 14 Constellations on right side of Milky Way and 14 Constellations on the left, while Abraham’s star-like descendants number 14 generation from Abraham to King David; 14 from David to the Babylonian exile and 14 from Babylon to Jesus.
Brahma and Sarasvati dwelt 100 years together then bore their first son, while Abraham was 100 when Sarah, at 90, bore Isaac .
and if you take the original semite abjad form it became even more aparent
Abraham brahma
brhm <brhm ITS IDENTICAL
lets review some part of the old testament
On that day, YHWH will punish
with his heavy, great and strong sword
the Leviathan, the Àeeing serpent,
yes, the Leviathan, the writhing serpent ±
he will kill the dragon in the sea.
leviatan is Early Indo-European influence of the Northwest Semitic world may have
come from various sources.17 Two of these are especially relevant: the
Anatolian Indo-European speaking cultures of Asia Minor (especially Hittites and Luwians) on the one hand, and Indo-Aryan inÀuence (probably
mediated through the Mitannians, who while mainly being ethnically
and linguistically Hurrian had a clear Indo-Aryan inÀuence or stratum
in their onomastics and traditions as well
As an example of the possibly parallel dragon slaying mythology from
Vedic India, lets look at to this piece of text:
He who after slaying the serpent released the seven rivers,
he who drove out the cows that were held back by Vala,
he who between the two rocks gave birth to ¿re,
victorious one in battles ± he, O men, is Indra!
In both cases, we have tales of a young, stereotypically ‘manly storm god, using lightning
as his weapon (in Indras case, often identifed with his so-called 9DMUD,
compare with the lightning-shaped ‘cedar weapon used by Baal in the
Baal Cycle and depicted on the Baal au foudre stele) defeating a terrible
serpent. And just as Baal receives his weapons for destroying the sea god
Yamm (an ally of the sea serpent) from the craftsman god Kothar-waHasis, Indra receives his weapons from the Vedic craftsman god Tva
https://portal.research.lu.se/ws/files/21683942/Studies_in_Isaiah_offprint_Wikander.pdf
but what about the principal god itself? a storm god lay behind the earliest conceptualization of the Jewish god YHWH according to jewish historians
so YHWH was probably a relative of indo european indra thor or zeus
https://www.haaretz.com/archaeology/.premium.MAGAZINE-jewish-god-yahweh-originated-in-canaanite-vulcan-says-new-theory-1.5992072
and can genetics help us resolve the mistery?
yes it can ,ashkenazi levites alongside kohanites the jewish priestly class have one of the highest proportion of r1a halogroup in the world around 60% and their derivate is not from a recent eastern european source but much more ancient when indo iranians were conquering all the middle east around 1600 bc and now we have samples in megido with 25-40%steppe ancestry
https://en.wikipedia.org/wiki/Levite
there is no doubt that an indo european tribe like mitanni or hittites conquered or influenced heavily the proto hebrews .
@Strmil
Those steppe ancestries are of two different kinds, I don't think that they would have arrived together as one culture. One is like the group that would have directly come southward through the Caucasus (catacomb) while the other went through central Asia. Correct me if I am wrong, but isn't Catacomb more like Yamnaya than like Sintashta?
As far as the deities are concerned, I think it a little bit of the opposite. YHVH is like Baal- a local thunder god which makes sense since the locals are farmers and lightening helps fertilize the soil. The original EMBA steppe groups wouldn't have had this, they would have had a god of clear skies (Dyeus). A thunder god would be more important for the farmers, so the thunderer is probably a sign of neolithic European farmer influence. This would explain the tangential similarities: a thunderer deity- originating with the first farmers makes one path down into Levant with the Anatolian movement down south and another one through Europe into the eastern European steppe.
The MLBA thunder god would have also undergone a change in the origin story, since it would now involve the sky father (EMBA) and the earth mother (neolithic European), while I don't really recall this being the case for either the thunder gods of the Levant farmers or the European farmer traditions (that we are aware of).
@tlt
when sintashta and other related groups spread over the steppe replacing steppe emba they brought with them part of the sncestry of catacoumb or the other way around like was the case with poltavovska but you may be right
What im absolutely sure is that the similarities in the role of gods , masculine behaviour, heroic character slaying the serpent , lighning .... are too similar to have arise independently and have all the characteristic of " pure " proto indo european ethos .
@Srtmil
No, there were two main, distinct migrations of steppe peoples into the Near East.
One group went south through the Caucasus and the other via Central Asia.
The former was rich in R1b-M269 and its language is unknown. The latter was rich in R1a-Z93 and spoke Indo-Iranian languages.
And I'm not seeing any Catacomb ancestry in the latter. That is, no Catacomb ancestry in Andronovo.
Thanks
Finally got the southern levant dataset merged and ran qpAdm for Megiddo_MLBA_outlierfamily using harneys rotating model.
1. The boy is definitely derived for R1a1a1, have checked it myself using the egenstrat file data. Dont think there will be more info to drill down further.
2. Alalakh_MLBA does not work with any sources. Megiddo_MLBA works. This is not what G25 shows, so just noting this point.
3. Testing 3 models with rotating strategy: Megiddo_MLBA + Rus_Catacomb &
Megiddo_MLBA + Sintashta + Dzharkutan & Megiddo + Kubano_tersk_late
a. Model 1: Right pops = Rightall (my usual eurasian right pops) + Sintashta + Dazharkutan
Left pops:
Megiddo_MLBA_outlierfamily
Megiddo_MLBA: 73.7 +- 2.1
Russia_Steppe_Catacomb: 26.3 +- 2.1
p-value: 0.86
result file: https://pastebin.com/90KpQVA5
b: Model 2: Right pops = Rightall + Rus_catacomb
left pops:
Megiddo_MLBA_outlierfamily
Megiddo_MLBA: 61.9 +- 3.4
Russia_MLBA_Sintashta: 25.1 +- 3.4
Uzbekistan_BA_Dzharkutan1: 13 +- 3.6
p-value: 0.40
Result file https://pastebin.com/58WPET9b
c. Model 3: Rightpops = Rightall + Sintashta + Dzharkutan
left pops:
Megiddo_MLBA_outlierfamily
Megiddo_MLBA: 68.4 +- 2.7
Russia_North_Caucasus_MBA(KBD001 & 002): 31.6 +- 2.7
pvalue: 0.51
https://pastebin.com/K6bGRVBf
Model 1 is simpler 2 way model and has higher probability than model 2 which is a 3 way model, although both are plausible. Model 3 is also simpler than model 2, and just slightly more probable.
The PCA shows that these outliers fall on a straight cline between megiddo & north caucasus. And none of the BMAC outliers have the steppe heavy ancestry that is required as a source here.
PCA plot https://imgur.com/Nonc2Fu
Ariel: I see a lot of people treating Mitanni as if it were synonymous with an Indo-Aryan ethnicity in the Middle East. This is a completely misconstrued assumption. The people and the kings of Mitanni all called themselves Hurrian. They called their homeland the Hurri-land, but the kingdom centered on the city of Wassukanni and ruled by the dynasty of Saustatar was called Mitanni. Once Wassukanni was utterly destroyed by the Assyrians the name Mitanni vanished.
Prior to the formation of Mitanni there already existed Hurrian-related people (Turukkians) mingled with Indo-Aryanized tribes in Iran. They were raiders warriors and moved around the Near East as far as south as Canaan, where we find some governors (at service of the Pharoh) with Indo-Aryan names. Their military order was heavily influenced by Indo-Aryan culture (no doubt due to the war chariot), but the people themselves spoke Hurrian, worshipped Hurrian gods and were nothing else but Hurrians to the other people of the Middle East.
Typically the ruling elite would put their own gods at the head of the pantheon. The supreme god of Mitanni was always Teshub, a native Hurrian god. The Indo-Aryan gods would come after Sumerian gods in their lists. Hinting rather than Indo-Aryans conquering the Hurrians, the Hurrians absorbed a segment of Indo-Aryan people and culture, prized for their military technology and abilities. They were born with Hurrian names but would adopt Indo-Aryan or Indo-Aryanized names if they took a military position.
@vAsiSTha
The R1a means that model 2 is much more plausible, even though it's a three-way model.
Also, the G25 clearly shows an Andronovo signal in this Megiddo outlier.
The G25 is better at picking up more realistic, proximate ancestry than qpAdm, which only focuses on deep ancestry proportions.
@vAsiSTha
"Finally got the southern levant dataset merged and ran qpAdm for Megiddo_MLBA_outlierfamily using harneys rotating model."
Stop throwing in wrong models. You don't understand qpAdm at all, you don't understand what:
1. All right populations should be at approximately the same distance from the left populations and the further the better.
2. qpAdm works based on f4 statistics, but when you use the right population that is in the left population you actually use f3 like statistics and they are not compatible.
3. It is absolutely impossible to compare models with different right populations, their coefficients and p-values are completely incomparable.
4. It is impossible to adjust the right population to your goals.
@Leron
Ariel: I see a lot of people treating Mitanni as if it were synonymous with an Indo-Aryan ethnicity in the Middle East.
No, that's just your impression.
We're actually talking about the subset of the Mitanni who were of Indo-Aryan origin.
@davidski
"The G25 is better at picking up more realistic, proximate ancestry than qpAdm, which only focuses on deep ancestry proportions."
thats bullshit. G25 loses a chunk of the information by condensing data into 25 coordinates. MBs of data is converted into a few bytes, how can it be more accurate?
@vAsiSTha
It's more accurate for proximate ancestry because it picks up recent drift, which qpAdm misses.
Nice work Vashishta but the time frame is not Yamnaya , R1a effectively guarantees it is of Steppe MLBA origin , in addition , the models include more late BMAC source or a Maykop Steppe enriched Andronovo source. They plot with Steppe rich Western Iranien like Azeri and Ezid for this reason.So I agree with David here.
Kouros said...
"Nice work Vashishta"
These are completely erroneous works, erroneous fitting under erroneous results, and falsification of results by parameter fitting.
@kouros
Rus_Catacomb is dated 2300bce. The PC steppe catacomb culture persisted till 1700bce as per wiki.
Kubano_tersk_late is dated 2000bce. These are not Yamnaya period.
Catacomb burial is found in Ukraine till 1100bce. with sample MJ09 from Ukraine which has very similar ancestry to North caucasus. So this ancestry persisted for long.
Target: Levant_Megiddo_MLBA_o1:I2200
Distance: 2.2777% / 0.02277661
69.6 Levant_Megiddo_MLBA
30.4 UKR_Catacomb
Not saying this is what happened as this sample is very late, but you get the gist.
There are no uniparentals from BMAC or andronovo yet (we dont know if R1a is indian L657, no need to jump the gun), but levant and near east abounds with R1b from PC steppe.
@ Archi
Time periode wise, these 3 outliers date to MLBA, I agree his first 1,3 models are strange , but its a good effort. I remember Kalash also model excellente with Yamnaya, so its also possible Andronovo enriched with some prior Yamnaya component though I doubt and its the reasons David mentioned. R1a shifts direction completely, so one has to use Andronovo related sources.
@ Vashishta
R1a absent even in those those late groups, archaeology disagree against this scenario absolument, this region to Beit Shamash show Mitanni influences, your model imagines this is Poland , Zlota people interact with the Yamna/early Corded Ware, except now Israelites. Scenario 1 never happened. R1a et chariote , in Megiddo is nothing but le Mittani.
Kouros said...
"Time periode wise, these 3 outliers date to MLBA, I agree his first 1,3 models are strange , but its a good effort."
These models are inherently erroneous because they do not fit the purpose. The goal of qpAdm is to estimate the proportion of ingroup source in the target population, but Megiddo_MLBA_o1 and Megiddo_MLBA_o2 do not make up a single population, there is no "Megiddo_MLBA_outlierfamily", this is a fact. Combining them in a Megiddo_MLBA_outlierfamily is a pure deception, and the results are not clear to what, the average hospital temperature.
And the use of ingroup sources in the right populations of the outgroup, it's actually something with something. It's also funny endless increase in the number of right populations, as if it gives something, the more the number of right populations the more likely they are to form a group and all the results are wrong, so it is enough 3-6 right populations, it no longer means better, it's even worse, it is much more important the quality of the right populations themselves.
@kouros
"this region to Beit Shamash show Mitanni influences, your model imagines this is Poland"
lol we are talking about a couple of juvenile outliers, not some ancient war story. they can be from anywhere. Let the data tell you where they are from, not your bias.
@archi
"there is no "Megiddo_MLBA_outlierfamily", this is a fact. Combining them in a Megiddo_MLBA_outlierfamily is a pure deception"
Kindly stop vomiting here.
I2200 and I2189 are siblings and labeled as Family4 in the supplementary excel file. I have analyzed these 2 samples, not the 3rd one who is not a sibling.
Have the decency to check my work which clear mentions no of samples as 2 before mouthing off like usual without an iota of brainpower behind it.
@ Archie, qpAdm is plagued with some issues like that that is important to look at the archaeology of the region.
@ Vashishta , catacomb burials at 1100 BC , even some Andronovo graves use red ochre powder like Yamnaya , continuation of tradition, not actual culture. The Mittani femme and the R1a in this child from the same time period point to movements from Asie Centrale.Also while going through posts, in one post the Steppe,BMAC amounts are same as in Indiens and they too have plenty R1a. The Steppe source in that Mittani femme and this outlier are same as in Soouth Asien groups it is of Andronovo related origin not Yamna , all are from same time period. It looks like one group of Indo Aryans migrate to West Asia other to South Asia around same time, but Indo Aryans in South Asia more successful because they had no competition and thats why r1a is so common populace there.
Even davidski has modeled Levant LBA with yamnaya earlier. The north caucasus, pc steppe connections are clear.
https://eurogenes.blogspot.com/2019/04/r1b-m269-in-bronze-age-levant.html
@Davidski
Any ideas when the bam files for the Megiddo R1a sample are released? I really want to know his R1a clade could be the first L657 in ancient dna.
@kouros
The R1a in this is not proven to be from south asia. it is what you wish, but it is not what is truth as of now.
The raw files will be uploaded here, but not done yet
https://www.ebi.ac.uk/ena/data/search?query=PRJEB37057
That R1a isn't from South Asia, it's from Central Asia, and Eastern Europe before that.
Think Abashevo > Sintashta-Petrovka > Andronovo > Indo-Iranians > Indo-Aryans > Mitanni
@ Vashishta , not biased but your modelling is very strange for this time. Based of the posts I read you jeer at David but he is not wrong here, he is telling the truth. R1a in the Western Asia is due to Indo Iraniens and they do exist in plenty numbers today in West Asia with R1a.
@kouros
No the modeling is not strange for the time at all. because look at this simple 2 way model using 1100bce Ukr_catacomb sample. literally any north caucasus and south steppe sample from this time period ie 2nd mill bce will work, G25 also shows the same.
left pops:
Megiddo_MLBA_outlierfamily
Megiddo_MLBA: 42.7 +-3.1
Ukraine_BA_Catacomb.SG: 57.3 +- 3.1
pvalue: 0.584
https://pastebin.com/QugwfXgH
snps of the ukr sample is low at 170k, but std errors are still tight. This is likely where the R1a came from.
I picked that Steppe shifted sample , this is not Europe, with Corded Ware in the Levant. Catacomb drops out as soon as Botai related sources added,
Target: Levant_Megiddo_MLBA_o2
Distance: 3.6766% / 0.03676650
40.8 Levant_Megiddo_IA
30.0 RUS_Sintashta_MLBA
21.4 UZB_Bustan_BA
7.8 KAZ_Kumsay_EBA
Target: Levant_Megiddo_MLBA_o2
Distance: 4.3214% / 0.04321447
59.0 Levant_Megiddo_IA
16.4 KAZ_Kumsay_EBA
16.0 RUS_Catacomb
8.6 RUS_Sintashta_MLBA
Target: Levant_Megiddo_MLBA_o2
Distance: 4.3602% / 0.04360185
60.2 Levant_Megiddo_IA
25.4 RUS_Catacomb
14.4 KAZ_Kumsay_EBA
Target: Levant_Megiddo_MLBA_o2
Distance: 4.5068% / 0.04506751
59.2 Levant_Megiddo_IA
34.4 RUS_Catacomb
6.4 KAZ_Botai
Target: Levant_Megiddo_MLBA_o2
Distance: 4.6956% / 0.04695602
57.0 Levant_Megiddo_IA
43.0 RUS_Catacomb
The outlier o2 , require additional Siberian/Botai/WSHG ancestrie which is what accompanied with BA Androvans/Indo Iraniens from Centrale Asie. The other 2 siblings not so much but follow same formation as this o2 girl. Siblings and o2 girl model well Steppe MLBA but 02 cannot be modeled with Catacomb due to all that Botai related stuff. David mentioned about the ancestral stuff and I see what he means. Your model does not reflect reality mon amie. Those R1b samples David posted from Levant have too low Steppe ancestry, as seen with Hittite . Steppe ancestry in Levant mainly from Indo Iraniens, additionally you have Indo Iranien people in Syria, Northern Iraq.
Target: Levant_Megiddo_MLBA_o1
Distance: 3.0420% / 0.03042030
60.8 SYR_Ebla_EMBA
20.6 RUS_Krasnoyarsk_MLBA
10.2 RUS_Catacomb
8.4 UZB_Bustan_BA
Target: Levant_Megiddo_MLBA_o1
Distance: 3.0923% / 0.03092289
56.8 SYR_Ebla_EMBA
30.6 RUS_Krasnoyarsk_MLBA
12.6 UZB_Bustan_BA
@ Coldmountains
"Any ideas when the bam files for the Megiddo R1a sample are released? I really want to know his R1a clade could be the first L657 in ancient dna".
In order to prove I2189 is an "Mitannian" Indo-Aryan, we only need to reach R1a-Z94,(R1a1a1b2a)in the subclades. It´s not necessary to go too far to R1a-L657 in the branch.
Read the article "post-Catacomb period in the Lower Volga region: from the krivoluk culture group to the Volga-don Babin culture" by R. A. Mimod Russian Academy of Sciences, Institute of archaeology.
Abstract. This article is devoted to an analysis of post-catacomb sites dating from the end of the Middle Bronze Age in the Lower Volga region and the Volga-Don interfluve. A description is provided of the funerary rite and the grave-goods accompanying the buri- als. It is argued that the pottery from the burials, which defines the culture, is represented by vessels decorated with multiple ridges, known also from the settlements of the region under discussion. On the basis of stratigraphic observations and cultural-typological com- parisons, synchronization lines of local post-catacomb antiquities are established, particu- larly in relation to the Dnieper-Don Babino and Lola Cultures. According to calibrated radiocarbon dates, the post-catacomb period in the lower reaches of the Volga was deter- mined as the 22nd–20th cc. BC. Examination of the terminology from a new angle has made it possible to conclude that the designation the so-called Krivaya Luka Cultural Group, which was previously used to mark post-catacomb sites in the region during the last ten years, has lost its relevance by this time. Given that the Lower Volga antiquities from the final stage of the Middle Bronze Age represent the eastern part of the Babino Cul- tural circle, a new term is suggested – «Volga-Don Babino culture» – in keeping with the principles used to denote other Babino cultures: Dnieper-Don and Dnieper-Prut Babino cultures.
Today, a significant part of the block of post - catacomb cultural formations is the Babino cultural circle. Its concept was formulated and convincingly developed by R. A. Litvinenko (2009; 2011B). It includes the Dnieper-don and of the Dnieper-Prut Babinski culture. There is no doubt that the post-catacomb monuments of the Volga-don interfluve and the Lower Volga Region, which for the last decade appeared in the archaeological literature under the name "krivolukskaya cultural group", represent the Eastern part of the cultural circle of Babino. It is advisable to assign the term "Volga-don Babin culture" to them, thereby integrating them into the taxonomy system proposed By R. A. Litvinenko, taking into account the use of toponyms of large rivers that delineate their territories as names for Babin cultures.
The attribution of the Volga-don and lower Volga post-catacomb materials to one of the Babin cultures also pays tribute to the historiographical tradition. One of the most striking complexes Fried Hillock 3/1, the study of which raised the question of the presence of post-catacomb antiquities on the Volga, S. Yu. Monakhov attributed it to the culture of polyvalent ceramics (now Babino) (Monakhov, 1984. P. 241-243). A.V. Kiyashko, one of the first to collect a representative selection of such burials, identified it as a local variant of the culture of polyvalent ceramics (Kiyashko, 2003.P. 31). At the present level of study of the cultural circle babineaux (of culture, of mnogovintovoy ceramic - CI) taxonomy, terminology, and its contents were seriously transformed, but the observations of scientists many years ago has not lost its relevance, and be blended into a new system of knowledge about pocketcomb by period Eastern Europe.
With the allocation of a new archaeological culture, the main structure of the Babino cultural circle was actually finalized, which now consists of three archaeological cultures: the Dnipro-Prut, Dnipro-don and Volga-don Babino cultures».
So it is quite possible that Indo-Europeans first penetrated into Central Asia and then Iran is not the Andronovo people, and the Babino culture. Since Andronovo and Sintashta originated apparently from Abashevo, they have an Eastern European component, such as GAC. Babino culture went to the South and they should not have this component.
In this case, Z2124 Ashkenazi may well be not Scythian, but Babino, that is, really the oldest and original. J1 they could meet in Iran, and there unite with them.
And their methods are similar to those of the Hindu Brahmans. High caste preachers from Z93 and plebs from other haplogroups
@Davidski
The genome-wide ancestry of the Megiddo outliers do not really need Central Asian ancestry when Northern Iran / East Anatolian populations included. And at this late time a Catacomb descendant group could have easily picked up some R1a even in Europe (from either Srubnaya or Andronovo), while their genome-wide ancestry did not necessarily even out.
So in this case, at that time I cannot see R1a as a decisive proof for Central Asian Andronovo origin.
Of course it is possible that such a Yamnaya-heavy group was also part of the Andronovo network in Central Asia, but then they would be different from core Indo-Aryans and their affiliation to them is anybody's guess.
@Srtmil
I would not draw such a detailed conclusion from general similarities between religious/mythological stories, but it is possible I guess.
@kouros
this is your best model for o2?
Target: Levant_Megiddo_MLBA_o2
Distance: 3.6766% / 0.03676650
40.8 Levant_Megiddo_IA
30.0 RUS_Sintashta_MLBA
21.4 UZB_Bustan_BA
7.8 KAZ_Kumsay_EBA
This is just my 2nd attempt on this outlier
Target: Levant_Megiddo_MLBA_o2:I10100
Distance: 2.6679% / 0.02667927
44.8 Levant_Megiddo_MLBA
29.2 ARM_MBA
21.6 RUS_Lola
4.4 RUS_Catacomb
Target: Levant_Megiddo_MLBA_o2:I10100
Distance: 3.2170% / 0.03216963
62.8 Levant_Megiddo_MLBA
37.2 RUS_Lola
this was the 1st attempt, still better than your model with sintashta and bmac.
@Slumbery
That's wishful thinking, especially considering that ALA019 is definitely from Central Asia and found nearby.
I do not have any particular wish about the matter, so wishful thinking is not the right word there even if I am completely off the mark.
Do you think that ALA019 cannot have a completely different origin than the Megiddo outliers? I see no reason why not.
The chances that an R1a in the Bronze Age Levant came via Catacomb rather than with Indo-Iranians from Central Asia are very slim, but the fact that we have indisputable evidence of someone directly from Central Asia already in the Bronze Age Levant makes them even slimmer.
@Davidski
Can you add Megiddo I10101 & I10270?
@Archi
Impossible. Not enough data.
Another simple two way model, with Armenia_MBA
Right pops: Rightall + Sintashta + Dzharkutan
left pops:
Megiddo_MLBA_outlierfamily
Megiddo_MLBA: 38.1 +-5.6
Armenia_MBA.SG: 61.9 +- 5.6
p-value: 0.5753
Result https://pastebin.com/7nbL18LU
A complicated model is not required at all
@vAsiSTha
A complicated model is not required at all.
It is if they have complicated ancestry, and they obviously do.
quote
"The two outliers from Megiddo (three including the sibling pair) provide additional evidence for the timing and origin of gene flow into the region.
These individuals are unlikely to be first generation migrants, as strontium isotope analysis on the two outlier siblings (I2189 and I2200) (Methods S1A) suggests that they were raised locally.
the only working northeast source population for these two individuals is the contemporaneous Armenia_MLBA, whereas the earlier Iran_ChL and Armenia_EBA do not fit "
The simplest models aren't necessarily the correct ones.
Armenia_MLBA just happens to work as the mixture source, because it manages to capture the complexity of the ancestry of these samples, but only to the limit of the resolution of this test.
Armenia_MLBA was practically after Megiddo outliers.
@Vladimir
>And their methods are similar to those of the Hindu Brahmans. High caste preachers from Z93 and plebs from other haplogroups
Basically this, mitanni conquered the proto hebrews and imposed thenselfs as priestly and warrior caste over the semite masses introducing that idea of supremacistic mesianism that was an aryan doctrine specially the indo aryan branch
judaism seem more and more like a very rudimentary version of hinduism mixed with local and mesopotamian traditions.
A curious thing isaac newton reached the same conclusion with completely different methods 300 years ago too , he thought proto hebrews ultimely came from india and it seem he have been proven right once again
Testing possible BA R1a sources for Megiddo_MLBA it can be noticed that I2189 prefers Corded Ware over Sintashta:
Target: Levant_Megiddo_MLBA_o1:I2189
Distance: 3.9937% / 0.03993725
33.2 IRN_Hajji_Firuz_C:I4351
21.4 Levant_Megiddo_MLBA:I10104
8.8 Corded_Ware_DEU:I1538
7.8 Levant_Megiddo_MLBA:I2190
7.6 RUS_Progress_En:PG2001
6.8 Corded_Ware_Baltic:Spiginas2
5.6 GRC_Mycenaean:I9006
3.4 RUS_Krasnoyarsk_MLBA:I1852
3.2 Corded_Ware_Proto-Unetice_POL:RISE431
2.0 Levant_Baqah_BA:I3986
0.2 RUS_Krasnoyarsk_MLBA:I1828
Target: Levant_Megiddo_MLBA_o1:I2200
Distance: 0.9614% / 0.00961443
37.0 IRN_Hajji_Firuz_C:I4351
9.2 RUS_Sintashta_MLBA:I1063
9.2 Levant_Megiddo_MLBA:I2195
7.6 RUS_Sintashta_MLBA:I1012
7.0 RUS_Progress_En:PG2004
7.0 Levant_Baqah_BA:I6566
6.8 Levant_Megiddo_MLBA:I10266
5.2 Levant_Baqah_BA:I3986
5.0 Corded_Ware_Proto-Unetice_POL:RISE431
4.4 Levant_Megiddo_MLBA:I4519
1.2 Levant_Megiddo_MLBA:I8187
0.2 Levant_Baqah_BA:I6462
0.2 GRC_Minoan_Lassithi:I0073
Target: Levant_Megiddo_MLBA_o2:I10100
Distance: 2.4295% / 0.02429467
26.8 Levant_Megiddo_MLBA:I2190
19.8 RUS_Progress_En:PG2001
16.6 IRN_Hajji_Firuz_C:I4349
12.6 Levant_Megiddo_MLBA:I10266
10.6 Corded_Ware_DEU:I0106
10.2 KAZ_Shoendykol_MLBA_Fedorovo:I10112
1.4 GRC_Minoan_Odigitria_low_res:I9131
1.0 RUS_Sintashta_MLBA:I1062
0.8 GRC_Minoan_Odigitria_low_res:I9129
0.2 Corded_Ware_DEU:I1536
Armenia_MBA.SG samples are dated between 1800-1200bce. Oldest being 1800bce. RISE413
RISE416
RISE423
So Archi, you are wrong as usual.
@davidski
"Armenia_MLBA just happens to work as the mixture source, because it manages to capture the complexity of the ancestry of these samples, but only to the limit of the resolution of this test."
There are 32 male samples from BMAC from 2200bce to 1400bce. Exactly 0 of them, including steppe mixed outliers, are R1a. Go figure.
Steppe_MLBA, whose descendants take over most of Central and South Asia during the metal ages, is 100% R1a.
Haha.
vAsiSTha Armenia_MBA.SG samples are dated between 1800-1200bce. you are wrong as usual.
You are mistaken as always, I wrote Armenia_MLBA, not Armenia_MBA. In Armenia_MBA only one sample is older than Megiddo outliers. That's why I've written practically/almost.
I've already written to you that the one who makes up his designations without explaining them to people is wrong, and even makes up wrong designations, if you call them then call them "outlier siblings".
Yes davidski, and if you combine both our statements you will find that the likely route to megiddo is straight from PC steppe, not through bmac.
Actually outlier siblings were probably already a couple of hundred years local in the Near East, they were born at all in Megiddo from an alliance with Megiddo woman, how they came from autosomes under such conditions is no longer understood. Autosomal traces do not last longer than 8 generations for marriage by local people, 200-250 years and everything dissolves in the local background.
Target: Levant_Megiddo_MLBA_o2:I10100
Distance: 1.7892% / 0.01789161
21.2 UZB_Sappali_Tepe_BA
18.0 KAZ_Zevakinskiy_LBA
14.6 Levant_Yehud_IBA
14.6 RUS_Afanasievo
11.0 IRN_Hajji_Firuz_C
9.0 TUR_Barcin_N
5.4 Levant_PPNB
4.4 Yamnaya_RUS_Caucasus
1.8 WHG
In spite of the fact that I2189 & I2200 are siblings it does not mean that their genetics coincide, as it is known, different set of chromosomes is passed to non twins.
Target: Levant_Megiddo_MLBA_o1:I2189
Distance: 2.6255% / 0.02625547
34.0 Levant_Yehud_IBA
20.0 RUS_Afanasievo
14.2 IRN_Hajji_Firuz_C
11.8 Levant_PPNB
10.4 UZB_Sappali_Tepe_BA
3.6 KAZ_Zevakinskiy_LBA
3.2 TUR_Barcin_N
1.8 RUS_AfontovaGora3
1.0 UKR_N
Target: Levant_Megiddo_MLBA_o1:I2200
Distance: 0.6871% / 0.00687117
20.8 ARM_Areni_C
20.0 Levant_Yehud_IBA
18.2 IRN_Hajji_Firuz_C
17.4 ARM_MBA
6.0 UKR_Meso
4.8 BGR_Beli_Breyag_EBA
4.6 UZB_Sappali_Tepe2_BA
3.4 UZB_Sappali_Tepe_BA
2.0 RUS_Afanasievo
1.8 Levant_Natufian
0.4 KAZ_Zevakinskiy_LBA
0.4 SRB_Iron_Gates_HG
0.2 ROU_Iron_Gates_HG
@Srtmil
So R1a-Z2122 is the Babino culture, and the modern Levites R1a-CTS6 are descendants of the mittani priests?
And R1b-Z2103 apparently passed from the steppe to Armenia and in General to Anatolia as part of Trialeti-Vanadzor culture
@Vladimir
"So R1a-Z2122 is the Babino culture"
Where did you get this from?
Weren't Greek Neolithics similar to Krepost?
"By contrast, data from five southern Greek Neolithic individuals (labelled ‘Peloponnese_Neolithic’)—three (plus one that has previously been published) from Diros Cave and one from Franchthi Cave—are not consistent with descent from the same source population as other European farmers. D statistics (Supplementary Table 2) show that these ‘Peloponnese Neolithic’ individuals, dated to around 4000bc, are shifted away from WHG, and towards CHG, relative to northwestern-Anatolian Neolithic and Balkan Neolithic individuals. We detect the same pattern in a single Neolithic individual from Krepost in present-day Bulgaria (I0679_d, 5718–5626bc)."
Interestingly enough Krepost is quite older.
Too many models are accepted for the Meggido outliers and they cannot all be true. This setup, even with outgroup rotation, may not have enough power to discriminate for recent drift for groups from Central Asia or the Caucasus. This can definitely happen. To clinch this we might have to try some modifications, or some other method.
@Archie
Where did you get this from?
it is clear that there is no data. So this is my hypothesis. I mean, it'd be logical
Here's more interesting. R. A. Mimohod. About the upper date of the volsko-lbishchenskaya cultural group, 2017. : "To date, it can be stated that the Genesis of cultural formations of the chariot block in the don region and the Volga region (Potapovka-Pokrovsk) involved, in addition to the main Abashevo component, cultural formations of the post-catacomb (Babin culture and krivoluk culture group) and more Northern, possibly post-cord (Volsk-lbishchevo cultural group and Voronezh culture) blocks."
How West Eurasian is the Botai or WSHG component?
Also Anatolian Turks and Iranian azeris look suprisingly Europid for being 16% and 8% East Eurasian on average respectivley.
Also interesting that the Eastern component doesn't really exist in Northern Anatolian Turks. Does that apply to most of the mountainous NE like Erzurum?
@Jatt,
When you ask how West Eurasian is Botai, this can be best answered relative to EHG. We generally separate ANE into West and East with EHG and WSHG, respectively.
So, firstly we can see that with respective to Kalash which is Southern, both EHG and WSHG are somewhat equidistant, this is the North-South divide, but we want to check this so we can separate out North-South drift (ANE vs South Asian) from East-West drift (EHG vs WSHG).
G25
Distance to: Kalash
0.27011209 KAZ_Botai:BOT14
0.28228454 RUS_Sidelkino_HG
And with f4
EHG Botai Kalash Chimp… -0.0056 0.00434 -1.30 22849 23108 489244
So we can see that there is no special drift/admixture between EHG and WSHG wrt to Kalash and can now look at the East-West difference between them.
Firstly, just looking at the distance them
Distance to: KAZ_Botai:BOT14
0.18754769 RUS_Sidelkino_HG
and wrt to MA1 and AG3
Distance to: RUS_MA1
0.10030768 KAZ_Botai:BOT14
0.15629714 RUS_Sidelkino_HG
Distance to: RUS_AfontovaGora3
0.10536370 KAZ_Botai:BOT14
0.16988152 RUS_Sidelkino_HG
So Sidelkino is further away from MA1 and AG3 compared to Botai. But is this an East-West cline or just drift specific to EHG? We check this by comparing wrt to Western Samples Iran_N, CHG and Anatolian HG.
Distance to: IRN_HotuIIIb_Meso:I1293
0.33085760 KAZ_Botai:BOT14
0.34905976 RUS_Sidelkino_HG
Distance to: GEO_CHG
0.36726873 RUS_Sidelkino_HG
0.37892660 KAZ_Botai:BOT14
EHG Botai CHG Chim… 0.0038 0.00614 0.612 22886 22714 489128
Not much of a difference with CHG and Iran_N, so we still dont know what makes EHG more Western than WSHG. We try with Anatolian HG.
Distance to: TUR_Pinarbasi_HG
0.39658587 RUS_Sidelkino_HG
0.46562506 KAZ_Botai:BOT14
EHG Botai Pinarbasi Chim… 0.0334 0.00663 5.04 18198 17022 375702
Ok, looks like we found something. Looks like we need to go way back and compare with Anatolian HG to get the Western position of EHG compared to WSHG.
@Davidski
"The chances that an R1a in the Bronze Age Levant came via Catacomb rather than with Indo-Iranians from Central Asia are very slim, but the fact that we have indisputable evidence of someone directly from Central Asia already in the Bronze Age Levant makes them even slimmer."
Well, the Megiddo outliers are nothing like ALA019, so I do not think the existence of ALA019 effects the probabilities much.
But otherwise you actually convinced me about the Central Asian route. Let's say Megiddo outliers are something like ((North Iran / Eastern Anatolia Calcolithic + Catacomb) + Andronovo) + Levant BA?
Something like this? (I also included ALA019, just to see the difference)
"sample": "Alalakh_MLBA_o:ALA019",
"distance": 1.9432,
"Gonur1_BA": 90,
"Yamnaya_RUS_Caucasus": 10,
"Hajji_Firuz_C": 0,
"Sintashta_MLBA": 0,
"Levant_Canaanite_MBA": 0,
"Arslantepe_LC": 0
"sample": "Levant_Megiddo_MLBA_o2:I10100",
"distance": 2.6181,
"Gonur1_BA": 20,
"Yamnaya_RUS_Caucasus": 14,
"Hajji_Firuz_C": 0,
"Sintashta_MLBA": 26.5,
"Levant_Canaanite_MBA": 22.5,
"Arslantepe_LC": 17
"sample": "Levant_Megiddo_MLBA_o1:I2200",
"distance": 0.9713,
"Gonur1_BA": 3.5,
"Yamnaya_RUS_Caucasus": 1.5,
"Hajji_Firuz_C": 36,
"Sintashta_MLBA": 27.5,
"Levant_Canaanite_MBA": 29,
"Arslantepe_LC": 2.5
"sample": "Levant_Megiddo_MLBA_o1:I2189",
"distance": 4.3478,
"Gonur1_BA": 0,
"Yamnaya_RUS_Caucasus": 13.5,
"Hajji_Firuz_C": 27.5,
"Sintashta_MLBA": 19,
"Levant_Canaanite_MBA": 8,
"Arslantepe_LC": 32
@Slumbery
"Well, the Megiddo outliers are nothing like ALA019,"
The fact is that they are all of different ages, the oldest of them is Megiddo_I10100 1688-1535 calBCE (3338±21 BP, RTK-7899), Alalakh ALA019 has an average age of 1625-1511 cal BCE, and Megiddo_I2189&I2200 are children who are buried later Megiddo_I10100. Unfortunately, I10101&I10270 have been ignored, which may be still Mitanni or Hittite children.
It should be taken into account that 330 Asian kings took part in the Battle of Megiddo.
I2163 2018 MathiesonNature2018 Pinhasi, Ron 1750-1625 calBCE Bulgaria_MLBA M R1a1a1b2
left pops:
Megiddo_MLBA_outlierfamily
Megiddo_MLBA: 70.3 +- 2.8
Bulgaria_MLBA: 29.7 +- 2.8
pvalue: 0.27
Result file https://pastebin.com/GBybRHR9
We should try to read the extended information about R subclades for I2189 and I10768 from BAMs https://www.ebi.ac.uk/ena/browser/view/PRJEB37057.
@Slumberry
"yHg J in EHG probably came from CHG and it is definitely not an ANE paternal lineage. CHG admixture of various levels was present in all EHG samples.
As for R being "EHG derived", the inflow of Siberian R lineages happened in the same time as EHG itself formed, so - even with likely later cross mixtures - more western populations received it in roughly the same time as EHG did"
My reply was poorly phrased you are right. But J could still have arisen in an ANE group, given that CHG and Iran_N are heavy in ANE ancestry or alternatively be specific to them.
I wouldn't call R Siberian per say but an ANE lineage. Siberians in my mind are more East Eurasian populations. But yes, you are correct, I only meant that Yamnaya and related populations got it from their EHG component.
I think that models for Megiddo_MLBA outliers are even better seen on PCA.
https://i.ibb.co/PYLQxsr/South-Levant-2020-new-PCA.png
You can see that earlier I10100 sample is exactly between Ancient Europe-North Central Asia border and Megiddo. Later samples of siblings are shifted more towards Megiddo.
I2163 2018 MathiesonNature2018 Pinhasi, Ron 1750-1625 calBCE Bulgaria_MLBA M R1a1a1b2
left pops:
Megiddo_MLBA_outlierfamily
Megiddo_MLBA: 70.3 +- 2.8
Bulgaria_MLBA: 29.7 +- 2.8
pvalue: 0.27
http://eurogenes.blogspot.com/2017/05/steppe-invaders-in-bronze-age-balkans.html?m=1
This only confirms my version that the first Indo-Europeans of the Levant are migrants from the Babino culture (Multi-cordoned Ware culture or Multiroller ceramics culture), with a fairly large substrate post katacomb or post Yamnaya, but with Y R1a-Z93+ . As you know, Babino expanded to the West (Noa culture, etc.), but also to the East.
“Quote Originally Posted by Pribislav View Post
I10768; 1600-1500 BC; Tel Megiddo, Israel; Megiddo_MLBA; R1b1a1a2-M269>L23>L51>L52
Although huge discovery if true, this should be taken with a grain of salt, since C>T and G>A transitions are the two most common types of aDNA damage, especially when covered with only one or a few reads.
M269 level: PF6441+ C>G (1G); CTS12478/PF6529+ G>A (1A)
L23 level: *no calls*
L51 level: Y410/E207+ G>A (1A)
L52 level: PF6544/CTS7650/FGC44/S1164+ C>T (2T)”
This situation is interesting. If L23 "no calls" could it mean that there is a chance that it is PF7562? And then it will really be like the Hittites
There was another M269 in Levant from an earlier study; with possible Italian or Aegean origin. At the same time there is East Med admixture in Iron Age Italy. So this might be a bidirectional flow between central and East Mediterranean
I2189 raw file is up. but only 48mb bam file. so i dont expect much resilution on y hg
The BAM file doesn't reveal anything beyond what the eigenstrat files already have. But at least we know it's R1a, so there's no mystery.
R1a arrived in the Levant from Central Asia with various Indo-Iranian groups, so we'll be seeing more of it in ancient DNA from the region. And there will be R1a-Z93 and even L657.
@Rob,
I agree with David that there were two Steppe routes into Middle East. Through Caucasus which brought R1b (including M269). And through Central Asia which brought R1a Z93.
So, the R1b M269 in Philistine has got to be from the Caucasus route ultimatly.
ACtually three I guess. The first was through Southeast Europe into Anatolia.
It is interesting, how even there were multiple Steppe IE migrations into Middle East they made a small impact on overall ancestry. Most Middle East pops have 5-10% Steppe ancestry, Iranians & Kurds have 13%.
L657 in BA levant? keep dreaming, you cant even find it in bmac or swat
If it ends up there in IA or later you will find it with aasi ancestry.
Early
Target: Levant_Megiddo_MLBA_o2:I10100
Distance: 1.4072% / 0.01407238
36.2 Levant_Megiddo_MLBA
16.2 RUS_Sintashta_MLBA
12.0 RUS_Afanasievo
11.8 ARM_MBA
10.6 UZB_Sappali_Tepe_BA
7.8 KAZ_Zevakinskiy_LBA
5.4 TUR_Barcin_N
Late
Target: Levant_Megiddo_MLBA_o1:I2189
Distance: 2.2812% / 0.02281223
52.2 Levant_Megiddo_MLBA
27.4 RUS_Afanasievo
10.8 TUR_Barcin_N
6.0 Levant_Yehud_IBA
3.6 UZB_Sappali_Tepe_BA
Target: Levant_Megiddo_MLBA_o1:I2200
Distance: 0.5754% / 0.00575449 | ADC: 0.25x
25.2 Levant_Megiddo_MLBA
23.6 ARM_Areni_C
18.4 ARM_MBA
16.8 IRN_Hajji_Firuz_C
12.2 RUS_Sintashta_MLBA
3.8 Levant_Yehud_IBA
@ Sam
“
So, the R1b M269 in Philistine has got to be from the Caucasus route ultimatly”
Yeah the z2103
But if there is any L151, then it’s from northwest
https://eurogenes.blogspot.com/2019/07/how-did-steppe-ancestry-spread-into.html?m=1
@vAsiSTha
Yeah L657 is from Indian tribals. Thanks for opening us the eyes.
L657 arrived in South Asia via the Gomal and Bolan pass. L657 is still high among South Pashtuns, Baluchs and Brahui from this region. The Khyber pass and Swat are not really so relevant here. Before this L657 was in BMAC and earlier in Andronovo>Abashevo/Fatyanovo>Sredny Stog. The big picture is already clear. Only the details need to be discussed.
@coldmountains
too bad you have 0 evidence for your hypothesis in the last 5 years. keep making up your own mythology
Youre never going to find L657 without some aasi ancestry
Haber 2020 samples from hellenistic and early roman period with 10% baloch type ancestry related to persian conquest. 0 R1a found from 4 male ssamples.
@vAsiSTha
Awesome arguments there.
@gamerz_J
But J could still have arisen in an ANE group, given that CHG and Iran_N are heavy in ANE ancestry or alternatively be specific to them.
I find it quite unlikely that J is an ANE lineage.
And I would not be sure that CHG or Iran_HG was rich on actual ANE ancestry either. They can be modelled with them, in the absence of better ancestral references, with abysmal fit.
Samuel,
Do you know of a good and accurate online calculator that will indicate an individual's proportion of 'Steppe' ancestry?
Unfortunately, the Gedmatch calculators don't seem to have a Steppe component per se.
Many thanks.
@gamerz_J
I wouldn't call R Siberian per say but an ANE lineage. Siberians in my mind are more East Eurasian populations."
Aren't the East Eurasian Siberians you are referencing (Ancient PaleoSiberians, First Peoples, NeoSiberians) later arrivals that superseded the Ancient North Siberians who were ancestral to ANE, and were they not West Eurasians?
@Matt
Thanks for the reply, I unfortunately only just noticed a few minutes ago , browsing through the paper now.
I am still not sure if it's CHG or Iran_N from Anatolia but looks more CHG-like.
@RobertN
Yes, though the Ancient North Siberians were not exactly West Eurasians as they had some additional ENA (East Asian affinities). But closer to West Eurasians than East Asians I think.
@Slumberry
"And I would not be sure that CHG or Iran_HG was rich on actual ANE ancestry either. They can be modelled with them, in the absence of better ancestral references, with abysmal fit."
I don't have that strong of an opinion on J, but what do you think Iran_N and CHG have if not ANE?
They show high affinity with EHG and MA1, do they not?
@robertN
IIRC ANE can be modeled as 75% gravettian and the rest ANA
If that is true it means EHG is the product of two different layers of WHG
the first gravettian ( proto WHG)
the second WHG formed after the LGM that is involved in the Sidelkino cluster
@old europe
Only it was the other way around. Now this way looks like this.
Altai, Karakol culture ->
Russian Plain, Streletskaya Culture ->
European Aurignac ->
Levantine Aurignac.
Gravetian culture came from Aurignac.
@old europe
" IIRC ANE can be modeled as 75% gravettian and the rest ANA
If that is true it means EHG is the product of two different layers of WHG
the first gravettian ( proto WHG)"
You're wrong.
Target: RUS_MA1:MA1
Distance: 8.4854% / 0.08485398
100.0 RUS_Yana_UP
0.0 CZE_Vestonice16
@John Thomas,
Eurogenes G25 PCA can get accurate Steppe ancestry scores. But, you have to use recent ancestor sources which makes it difficult. I have Steppe scores for modern pops. If you have your G25 coordinates, I can create scores for you.
@gamerz_J
"Yes, though the Ancient North Siberians were not exactly West Eurasians as they had some additional ENA (East Asian affinities). But closer to West Eurasians than East Asians I think."
I wasn't aware this was so controversial or disputed. In some articles it's stated as a given.
“'Remarkably, the Ancient North Siberians people are more closely related to Europeans than Asians and seem to have migrated all the way from Western Eurasia soon after the divergence between Europeans and Asians,' said co-lead author Professor Laurent Excoffier, from the University of Bern."
Are the East Asian affinities you mention a result of actual admixture, or are they reflective of a common recent ancestry with East Eurasians before they split? Maybe they were simply less drifted than other West Eurasians?
@gamerz_J
..., but what do you think Iran_N and CHG have if not ANE?
1. Common ancestry from a third, not yet sampled, ancient population.
2. As a variation of the above CHG/IHG had ancestry from a ANE subgroup that diverged from the known ANE before MA1.
Note that EHG "likes" AG3 much more than MA1, while CHG/IHG is the opposite. I remember some commentators claiming it is CHG admixture in MA1, but that is anachronistic.
Nevertheless, I am not sure either, but I do not see big amount of actual ANE ancestry in CHG/IHG as an obvious, settled matter.
@RobertN
"I wasn't aware this was so controversial or disputed. In some articles it's stated as a given"
Yes in most papers it is and I also think that they were more of a WE than ENA population,so I should have phrased this better. But, there are some scenarios where it might not be such a clear case.
"Are the East Asian affinities you mention a result of actual admixture, or are they reflective of a common recent ancestry with East Eurasians before they split? "
I am not sure, most studies seem to suggest admixture, but I certainly wouldn't exclude alternative explanations like those you mentioned.
@Slumberry
"1. Common ancestry from a third, not yet sampled, ancient population.
2. As a variation of the above CHG/IHG had ancestry from a ANE subgroup that diverged from the known ANE before MA1."
Interesting, both scenarios seem plausible to me. But there is something ANE-related , many Zagros samples had subclades of R.
"I remember some commentators claiming it is CHG admixture in MA1, but that is anachronistic."
Maybe I have asked you before, but do you think whatever ENA Iran_N had could be related to that of ANE? I am wondering if that could explain this affinity.
On the Megiddo outliers with steppe admxiture: The Bible mentions Hittites on some occasions. One is Uriah the Hittite, fighting in King Davids army, only to be sent to an untimely death because David fancied his wife. While this would make him probably far more recent than the Megiddo outliers it does goes to show that there is at least a possibility that this points to Hittites fighting for Jewish kings.
Off course is the bible not a historical account. There is barely some evidence that a king David even existed. But it does account for the fact that the Jewish authors considered Hittites fighting in Israel *conceivable*.
Iran_N (Zagros) and CHG (Kotias) do have ANE, but that ANE has nothing to do with yDNA J. I suspect that J came from the proto WHG in the Dzudzuana type of ancestry which they had.
@gamerz_J
"Maybe I have asked you before, but do you think whatever ENA Iran_N had could be related to that of ANE? I am wondering if that could explain this affinity."
The ENA in Iran_N from the Dzudzuana preprint list is Onge, the ENA in ANS has not yet been classified into an ENA branch as far as I know.
@Epoch
From Ezekiel 16:3
"And say, Thus saith the Lord GOD unto Jerusalem; Thy birth and thy nativity is of the land of Canaan; thy father was an Amorite, and thy mother an Hittite."
But it was pointed out once elsewhere that there was also a semitic tribe that was referred to as Hittite, so maybe that is what is being referenced above?
@Davidski,
Based on these admixture models, how far towards South Asia did the Mitanni swing on their way into their Hurrian settlement? It would be interesting to see their split or interaction
@Aniasi
Considering the lack of a South Asian signal in these samples, it looks like their ancestors came from somewhere northwest of Pakistan.
@TLT
"The ENA in Iran_N from the Dzudzuana preprint list is Onge, the ENA in ANS has not been classified into one of the ENA branches yet."
I am aware of this, but I am not convinced it is actually Onge(-like). I am not even convinced Onge is that good of a proxy for AASI ancestry (as VasiSTha has also pointed out) , albeit probably better than any other reference.
As you know, Iran_N is heavily basal, and thus that may complicate matters in finding a good model for its ancestry.
It is not unlikely that the ENA it has is related to the ENA in ANE, as even the Dzudzuana paper modeled CHG with Tianyuan instead, and it is seems improbable to me that the ENA in CHG is different from that in Iran_N.
@Davidski
Regarding ANE, this pre-print was just posted about the Salkhit individual: https://www.biorxiv.org/content/10.1101/2020.06.03.131995v1
I was wondering, they seem to model Yana as being 40% or more ENA, do you know/have any thoughts on how accurate that is?
In their alternative models it's less, but Tianyuan is 4% Goyet-like.
@gamerz_J
I don't have any interest in highly theoretical deep ancestry models such as these anymore.
They can change due to a lot of factors, and often it takes very little to shift the results significantly, so whatever.
@Davidski
Thanks for the reply, you do have a good point. This is probably the 5th one I see regarding the WE/ENA ancestry of Yana and related pops.
Global 25 said 70-30, I guess it's a good enough approximation for now.
@gamerz_J
"It is not unlikely that the ENA it has is related to the ENA in ANE, as even the Dzudzuana paper modeled CHG with Tianyuan instead, and it is seems improbable to me that the ENA in CHG is different from that in Iran_N."
The p-value for Tianyuan in CHG was worse than the p-value for Onge in Iran_N. Yes I know that Onge is a bad proxy but the Tianyuan is a worse one. Also, Tianyuan was autosomally rejected as a mix with Sunghir for ANS. The proper identity of both the west Eurasian (Sunghir rejected) and the east Eurasian (Tianyuan rejected) ancestors of ANS are still unknown so I wouldn't jump the gun for the identity of ENA in Iran_N and CHG especially with a worse p-value.
Concerning the comment:
"The Megiddo samples include a trio of interesting outliers dated to 1600-1500 BCE with significant ancestry from the steppe. One of these individuals is a male, I2189, who belongs to Y-haplogroup R and probably R1a. So he might also be of Indo-Aryan origin."
I asked one of the researchers about the Y Haplogroup R sample and she said stated:
"Hg R* stands for it's own. It means that only M-207 marker which defines this haplogroup is present but no other SNPs (which allow to define its derived haplogroups R1, R2 and etc.)"
So if their tests did not match for R-M420, then why assume otherwise (much as I would like you to be right as opposed to this sample being an R3 or R4)?
Just curious.
@Unknown
I2189 definitely belongs to R1 and probably R1a, for example...
R1 M640 DER
R1a1a1~ F3551 DER
More info here...
https://anthrogenica.com/showthread.php?20461-The-Genomic-History-of-the-Bronze-Age-Southern-Levant&p=671600&viewfull=1#post671600
Also, just because there are some R1 defining SNPs missing in the data doesn't mean this sample might actually belong to R3 or R4.
The data are missing because this sample is low quality. A higher quality version would have reads at those SNP sites.
And as per the genotype data, there are reads at other SNPs that show this individual to belong to R1.
@vAsiSTha
"Brahmins have lesser steppe than North and NW indians and pakistanis/afghans/kalash etc."
That's incorrect. There is substantial variation in NW India and Pakistan wrt steppe. In
India, the only group that clearly has higher steppe than Brahmins are the Hindu Jats and
Rors of Haryana, Western UP, Kalasha obviously. The Indo-Aryan speaking Brahmin groups by
and large have similar steppe like that of Khatris, Aroras, Gujjars, Potwaris, Brahuis,
Kohistanis, Balochis, Tarkhans, etc obviously with individual variations and obviously have
higher steppe than PJL like poeple and historically mid-low caste groups of the region. In
NW India (Punjab, Haryana, Rajasthan) the Brahmins are at the very top when steppe
ancestry is concerned, except for Jats. Punjabi Jatts themselves have quite a variation in
steppe scores, making the average not far ahead of the Brahmins unlike Haryana UP Jats and
Rors who are significantly ahead of all other S Asian groups. There are Nepali Brahmin
samples having very high steppe. Outside of NW (Punjab, Haryana, Rajasthan, western UP),
Brahmins have significantly high steppe than the other groups of the region, including
Dravidian speaking Brahmins.
Post a Comment