PDA

View Full Version : aDNA from Kumptepe (LATE Neolithic Anatolia)



Gravetto-Danubian
01-02-2016, 12:28 AM
In 2015, we saw posted abstracts giving us a teaser about Late Neolithic genomes from Anatolia, Kumptepe (6700 y BP, 4 700 BC)

The original abstract caused some confusion as to what it mean by its affinities to Caucasus populations ? Ie was there a genetic shift in the Late Neolithic and appearance of a "Caucasus' like component not seen in earlier Farmers ? Was it the same "Teal' seen in Yamnaya -type groups ??

The full paper now available (or at least coming imminently)

http://www.cell.com/current-biology/abstract/S0960-9822%2815%2901516-X



Highlights


•Kum6 shows a strong population continuity with present-day Sardinia
•Kum6 expresses connections to the central Eurasian gene pool
•Kum6 shares notable affinity with the Iceman, a 5,300-year-old southern European
•Genetic affinities to both East and West suggest continuous contact with Anatolia



Summary

Anatolia and the Near East have long been recognized as the epicenter of the Neolithic expansion through archaeological evidence. Recent archaeogenetic studies on Neolithic European human remains have shown that the Neolithic expansion in Europe was driven westward and northward by migration from a supposed Near Eastern origin [ 1–5 ]. However, this expansion and the establishment of numerous culture complexes in the Aegean and Balkans did not occur until 8,500 before present (BP), over 2,000 years after the initial settlements in the Neolithic core area [ 6–9 ]. We present ancient genome-wide sequence data from 6,700-year-old human remains excavated from a Neolithic context in Kumtepe, located in northwestern Anatolia near the well-known (and younger) site Troy [ 10 ]. Kumtepe is one of the settlements that emerged around 7,000 BP, after the initial expansion wave brought Neolithic practices to Europe. We show that this individual displays genetic similarities to the early European Neolithic gene pool and modern-day Sardinians, as well as a genetic affinity to modern-day populations from the Near East and the Caucasus. Furthermore, modern-day Anatolians carry signatures of several admixture events from different populations that have diluted this early Neolithic farmer component, explaining why modern-day Sardinian populations, instead of modern-day Anatolian populations, are genetically more similar to the people that drove the Neolithic expansion into Europe. Anatolia’s central geographic location appears to have served as a connecting point, allowing a complex contact network with other areas of the Near East and Europe throughout, and after, the Neolithic.

Passa
01-02-2016, 12:34 AM
I tried accessing full text both by laptop and mobile phone, but a "system problem" message appears.

Gravetto-Danubian
01-02-2016, 12:35 AM
I tried accessing full text both by laptop and mobile phone, but a "system problem" message appears.

Same here. They must still be loading it onto system. Try again later.

ADW_1981
01-02-2016, 12:35 AM
In 2015, we saw posted abstracts giving us a teaser about Late Neolithic genomes from Anatolia, Kumptepe

The abstract originally suggested genetic flux in late Neolithic compared to early Neolithic, and appearance of a "Caucasus' like component not see in earlier Farmers. What did this mean, everyone thought ? Was it the same "Teal' from Yamnaya -type groups ?

Full paper now available

http://www.cell.com/current-biology/abstract/S0960-9822%2815%2901516-X

I'm not certain what that meant, and I may have missed it. I'm pretty certain Sardinians when used as a proxy to the Neolithic farmers have little to no resemblance to CHG, let alone the related Teal in Yamnaya.

Leeroy Jenkins
01-02-2016, 12:41 AM
If someone here has access, can you tell me what to make of this super blurry image I cropped, or better yet, post a copy of better quality?

http://i67.tinypic.com/331lz43.png

The disappearance of the green and orange components and the introduction of the light blue component during the Chalcolithic period is interesting, but I am not sure what the components represent nor which populations these samples were taken from due to the low quality. I assume the sample with the red font is the Kumtepe sample, but even that is just speculation. :(

ADW_1981
01-02-2016, 12:47 AM
Thanks, the only one I can tell with certainty is Dark Blue = Euro Hunter Gatherer, not sure if it's WHG + EHG inclusive or separated. Judging by the Pie graphs I think it's inclusive of both east and west hunters.

Gravetto-Danubian
01-02-2016, 01:05 AM
I Have access

I'll post useful tidbits


SAMPLE:

We generated genome-wide sequence data from two Neolithic individuals excavated at the site of Kumtepe and obtained 0.133 genome coverage for Kum6 (6,700 BP) and
0.013 for Kum4 (5,500–4,800 BP)


RESULTS:

* Kum6 carries the H2a3 mitochondrial haplogroup (both females, the other Kum4 was not typed)

* The Anatolian Kum6 individual falls close to the early and middle Neolithic European farmers, showing a tendency toward modern-day Near Eastern populations. Interestingly, Kum6 does not group with any modern-day Anatolian populations. These results were confirmed by outgroup f3 statistics where, among modern-day groups, Kum6 shows the greatest genetic similarity to Sardinians, Greeks, and Cypriots, whereas modern-day Anatolian populations display lower levels of genetic affinity to Kum6 (Figure 2). Kum6 also falls between modern-day West Asians and Europeans when additional modern-day populations are included in the analysis

* For a model with nine clusters (K = 9; results for higher numbers of clusters are similar, Figure S3), three major ancestry components were observed in the ancient individuals. The first one (blue), observed as the main component in all hunter-gatherers, is also found as a minor contribution to all farmers, which is in line with the observed admixture from hunter-gatherers into farmers [3]. The second (orange) and the third (green) components were observed mostly in farmers to varying degrees (5%–68% and 0.06%–45% for K = 9, respectively). The orange component is mainly found in present-day Western Europeans, whereas the third component (green) is mostly found in the modern-day Near East and Caucasus, and the highest proportion of this third component among Neolithic individuals was observed in Kum6 (45% for K = 9). The notion that this component is West Asian is also supported by its presence in a Bronze Age Armenian sample (51%), which contains less than 2% of the orange component. Interestingly, this ‘‘West Asian’’ component (green) is not related to the potential genetic material brought to Europe by migration during the Bronze Age and recently connected to the Yamnaya culture [19, 24], visualized in Figure 3 as light blue, and it is observed in high frequency in modern-day people from southern Asia. The elevated ‘‘West Asian’’ affinity of Kum6 is likely to be the cause of the genetic differentiation observed between Kum6 and all other ancient farmers shown in the PCA plot (Figure 1B.).

* A clear decline was observed in the values of the green component over time (average of _29% in Early Neolithic, _14% in Middle Neolithic, and 2% in Late Neolithic), which is consistent with increased admixture with hunter-gatherer groups [3]. Our results suggest that the two ancestry components of ancient farmers (orange and green in Figure 3) were established at an early stage, probably before the first farmers expanded into Europe, and were maintained in Europe up until the end of Middle Neolithic and that both components are present in various modern-day European populations. Therefore, these observations directly link the early European Neolithic gene pool to western Anatolia.

* We computed D statistics [20] to further investigate additional genetic relationships between ancient Europeans with
sufficient sequencing coverage (>13) and Kum6. All proposed tree topologies where the Tyrolean Iceman [20] was included as one of the in-groups were rejected (2 < jZj < 4.6), suggesting gene flow or a more recent shared ancestry between Kum6 and the Tyrolean Iceman (Figure 4A). A similar tendency was observed with a Middle Neolithic Hungarian farmer [23], (co1), contemporary with the Tyrolean Iceman, resolution due to the low coverage of Kum6 and Co1. The observed genetic affinity between the Tyrolean Iceman and Kum6 could be interpreted as additional contacts between western Anatolia and Neolithic Europe at a later stage. This scenario is congruent with mitochondrial [29] and archaeozoological [30] studies, as well as the archaeological indications of multiple waves of contact between the Balkans and Anatolia.

* Furthermore, the Bronze Age Yamnayan component suggested to be a part of the Corded ware expansion [19, 24] is
not present in Kum6, and thus is not producing any increased affinity to the ancestors of the Yamnaya culture from north of the Caucasus (D-Denisovan, Yamnaya_RISE; Kum6, early farmer), all Z > _1.7).

* Contacts to the east, independent of Yamnaya ancestry [19, 24] are, however, supported by (1) higher affinity
of Kum6 to some Bronze Age Asian cultures when compared to Mesolithic Europeans and (2) higher affinities of Bronze Age Asians to Kum6 compared to early Neolithic Europeans (Table S5).


* "A comparison of Kum6 to an Asian Upper Paleolithic individual (Ust-Ishim [27]) and a European Upper Paleolithic sample (Kostenki14 [28]) confirms that Kum6 shows more affinity to early Europeans (Z = 5). Stronger affinities of Kostenki14 to Kum6 than to early Neolithic Europeans (Table S5), however, suggest that Kum6 contains genomic components not found in early Neolithic Europeans. Kum6 is also grouped with farmers in a model-based
population-tree analysis [34] (Figure 4B.), and the inferred migration edges point to the same conclusion as the D statistics results, as well as manifest the expected signals from previously published observations [3, 4].

* "Interestingly, the genetic similarity to the Tyrolean Iceman and the eastern component detected in Kumtepe indicate an intense entanglement of contacts from the East and into Europe, with western Anatolia at the center. Most modern-day European populations display ancestries from Mesolithic hunter-gatherers, early Neolithic farmers, and in some cases traces of additional admixture from different sources [3, 4, 23, 24] (Figure 3). Modern-day Anatolian groups display a variety of admixture traces originating from groups in the Middle East, Central Asia, and Siberia, which
cause Kum6 to be genetically more similar to modern-day Europeans than to modern-day Anatolians.

Leeroy Jenkins
01-02-2016, 01:10 AM
Interestingly, this ‘‘West Asian’’ component (green) is not related to the potential genetic material brought to Europe by migration during the Bronze Age and recently connected to the Yamnaya culture [19, 24], visualized in Figure 3 as light blue, and it is observed in high frequency in modern-day people from southern Asia. The elevated ‘‘West Asian’’ affinity of Kum6 is likely to be the cause of the genetic differentiation observed between Kum6 and all other ancient farmers shown in the PCA plot (Figure 1.

Interesting. Does this add weight to the belief of some that the non-EHG ancestors of Yamnaya peoples were from Central Asia and not from the Caucasus?

Gravetto-Danubian
01-02-2016, 01:12 AM
delete

ADW_1981
01-02-2016, 01:15 AM
I Have access

I'll post useful tidbits

The orange component is mainly found in present-day Western Europeans, whereas the third component (green) is mostly found in the modern-day Near East and Caucasus, and the highest proportion of this third component among Neolithic individuals was observed in Kum6 (45% for K = 9). The notion that this component is West Asian is also supported by its presence in a Bronze Age Armenian sample (51%), which contains less than 2% of the orange component. Interestingly, this ‘‘West Asian’’ component (green) is not related to the potential genetic material brought to Europe by migration during the Bronze Age and recently connected to the Yamnaya culture [19, 24], visualized in Figure 3 as light blue, and it is observed in high frequency in modern-day people from southern Asia. The elevated ‘‘West Asian’’ affinity of Kum6 is likely to be the cause of the genetic differentiation observed between Kum6 and all other ancient farmers shown in the PCA plot (Figure 1B.).

Very cool, but for some reason I am not seeing any Light Blue in the west Europeans. I would still expect to see a little bit. It might be the blurriness that is causing this.

Gravetto-Danubian
01-02-2016, 01:18 AM
7140

7139

7141

Gravetto-Danubian
01-02-2016, 01:34 AM
Moderns and ancients admixture from Supp:

7142

7143

Sorry, i'll post less hazy version from home later today

royking
01-02-2016, 01:38 AM
Looks quite good. Kum6 has much CHG-like admixture and situates actually close to Cyprus.

J Man
01-02-2016, 02:10 AM
Kum6 belongs to mtDNA haplogroup H2a3 that is interesting.

Chad Rohlfsen
01-02-2016, 02:17 AM
No, that does not look like CHG. It's not something present in Yamnaya, from what I see. When looking at lower K's, Kum6 looks like other EN samples. There's no EHG and CHG components in this. None of the components in Yamnaya at lower K's are in Kumtepe samples. The coverage of these genomes is rather poor too. The stats showing no increased affinity to Yamnaya, over other EN, kind of proves the point. If they have any CHG, it's so minor as to not make them really any closer to Yamnaya. Sorry to burst any bubbles here, but there will not be squat for CHG in SE Europe before the metal ages, and certainly could be lacking completely in CT, before anyone goes there again.

royking
01-02-2016, 02:27 AM
Sorry Chad,
I disagree--Look at K=12. I don't have the blinders on about Yamnaya which interests me little. One can readily see the CHG shift in Kum6 at K=12.

Chad Rohlfsen
01-02-2016, 02:38 AM
Roy,

You need to look at stats and not some bullshit, K12 modern cluster stuff. It's nothing present in Yamnaya. That k12, orange component is Middle Neolithic, not EN. That green is an increased NE affinity, also present in other EN samples, which will be almost identical in Neolithic Anatolians, if they tested them. This Kumptepe sample will probably be identical to Anatolians. Not only that, the quality is so low, I wouldn't even bother with these samples. You guys put too much weight on Admixture runs in these papers, which is always based on modern clusters, and not ancients. You need to look at stats!

Gravetto-Danubian
01-02-2016, 02:41 AM
Roy,

That k12, orange component is Middle Neolithic, not EN. That green is an increased NE affinity, also present in other EN samples, which will be almost identical in Neolithic Anatolians, if they tested them. This Kumptepe sample will probably be identical to Anatolians. Not only that, the quality is so low, I wouldn't even bother with these samples. You guys put too much weight on Admixture runs in these papers, which is always based on modern clusters, and not ancients. You need to look at stats!

The paper methodology section admits to poor coverage, but suggests ""Genome coverage as low as 0.133 has previously been shown to provide important and clear insights into population history""

So it'll be interesting what some of the clever stat-minded contributors make of this extra, admittedly lone sample.

Moderator
01-02-2016, 03:07 AM
Roy,

You need to look at stats and not some bullshit...

Please be civil in your discourse with other members. Thank you for your cooperation.

Chad Rohlfsen
01-02-2016, 03:33 AM
Moderator,
To be clear, I was calling the Admixture output bullshit, and not the poster's comments.

Further, here is something posted by Gravetto-Danubian..

* Furthermore, the Bronze Age Yamnayan component suggested to be a part of the Corded ware expansion [19, 24] is
not present in Kum6, and thus is not producing any increased affinity to the ancestors of the Yamnaya culture from north of the Caucasus (D-Denisovan, Yamnaya_RISE; Kum6, early farmer), all Z > _1.7).

Then, look here.

Denisovan Yamnaya_Kalmykia Anatolia_Neolithic LBK_EN 0.0044 1.746 23450 23246 492005

Kumtepe looks like an Anatolian to me.


Edit: Please, refer to stats, rather than Admixture output based on modern components. Especially, at the higher K's. They will lead you astray, very fast.

vettor
01-02-2016, 04:30 AM
The orange component is mainly found in present-day Western Europeans, whereas the third component (green) is mostly found in the modern-day Near East and Caucasus, and the highest proportion of this third component among Neolithic individuals was observed in Kum6 (45% for K = 9). The notion that this component is West Asian is also supported by its presence in a Bronze Age Armenian sample (51%), which contains less than 2% of the orange component. Interestingly, this ‘‘West Asian’’ component (green) is not related to the potential genetic material brought to Europe by migration during the Bronze Age and recently connected to the Yamnaya culture [19, 24], visualized in Figure 3 as light blue, and it is observed in high frequency in modern-day people from southern Asia. The elevated ‘‘West Asian’’ affinity of Kum6 is likely to be the cause of the genetic differentiation observed between Kum6 and all other ancient farmers shown in the PCA plot (Figure 1B.).

Very cool, but for some reason I am not seeing any Light Blue in the west Europeans. I would still expect to see a little bit. It might be the blurriness that is causing this.

orange represented EEF in the various 2015 papers, the dark green appears the Levant/ME , dark Blue looks like WHG and sky blue is south Asian

Arame
01-02-2016, 05:09 AM
Working direct links to Supplemental PDFs.

http://www.cell.com/cms/attachment/2042476822/2055464037/mmc1.pdf

http://www.cell.com/cms/attachment/2042476822/2055464038/mmc2.pdf

Arame
01-02-2016, 05:51 AM
Sky blue is the Gedrosia? That is present in Yamna.

Generalissimo
01-02-2016, 06:25 AM
Sky blue is the Gedrosia? That is present in Yamna.

And it's irrelevant.

Krefter
01-02-2016, 06:35 AM
Looks like Chad is right. Although in D-stats they show a bunch of Steppe groups are closer to Kum6 than other EEFs. Armenia_BA is most significantly closer to Kum6 than to other EEFs. Kum6 is from 6700 BP, so not much younger than Stuttgart. There's an about 6,000 year old genome from Greece that's EEF. CHG could and other elements have arrived by 3000 BC. We can't decisively say CHG or anything non-EEF didn't arrive in SE Europe till the Metal ages.

It's cool narrative. A close-knit group colonized most of Europe, and from 5000-3000 BC pretty much everyone in Europe and Turkey was apart of this group. Then starting by at least 3000 BC, people from the Steps and West Asia moved into their territory and ended their reign. And Sardinians are the last almost pure ones let. Without aDNA it would be appear the distinctiveness of Sardinians is just because of isolation, but they represent people that took up most of Europe and Turkey for 1,000s of years.

After Stuttgart, Loschbour, Otzei, etc. it was a mystery how North Europe became non-EEF. Because of lots of DNA from east of Germany we know it's Steppe. But now we need DNA from SE Europe and West Asia dating 6,000-3,000yo to find out who mixed with EEF there. Some were Steppe, but the main source came from West Asia. BTW, Generalissimo I don't think it came in historical times. Italy has the lowest IBD with each other for any European country and has lowest IBD with other Europeans. If West Asian or Balkan admixture came in historical times Italians should have more IBD with each other and or West Asia/Balkans. I think it came in sometime between 0 AD and 3000 BC.

Gravetto-Danubian
01-02-2016, 06:57 AM
Looks like Chad is right. Although in D-stats they show a bunch of Steppe groups are closer to Kum6 than other EEFs. Armenia_BA is most significantly closer to Kum6 than to other EEFs. Kum6 is from 6700 BP, so not much younger than Stuttgart. There's an about 6,000 year old genome from Greece that's EEF. CHG could and other elements have arrived by 3000 BC. We can't decisively say CHG or anything non-EEF didn't arrive in SE Europe till the Metal ages.

It's cool narrative. A close-knit group colonized most of Europe, and from 5000-3000 BC pretty much everyone in Europe and Turkey was apart of this group. Then starting by at least 3000 BC, people from the Steps and West Asia moved into their territory and ended their reign. And Sardinians are the last almost pure ones let. Without aDNA it would be appear the distinctiveness of Sardinians is just because of isolation, but they represent people that took up most of Europe and Turkey for 1,000s of years.

After Stuttgart, Loschbour, Otzei, etc. it was a mystery how North Europe became non-EEF. Because of lots of DNA from east of Germany we know it's Steppe. But now we need DNA from SE Europe and West Asia dating 6,000-3,000yo to find out who mixed with EEF there. Some were Steppe, but the main source came from West Asia. BTW, Generalissimo I don't think it came in historical times. Italy has the lowest IBD with each other for any European country and has lowest IBD with other Europeans. If West Asian or Balkan admixture came in historical times Italians should have more IBD with each other and or West Asia/Balkans. I think it came in sometime between 0 AD and 3000 BC.

It came after c. 4500-3500 BC

Kumtepe is "the last of his her kind", or the first trace of a new kind (?)

Arame
01-02-2016, 08:31 AM
There is no comparison with Kotias/Satsurbia? So this paper is written before CHG was known ?

rozenfeld
01-02-2016, 08:41 AM
There is no comparison with Kotias/Satsurbia? So this paper is written before CHG was known ?

Mostly yes:

This paper:
Accepted: December 9, 2015
Received in revised form: December 7, 2015
Received: November 20, 2015

Paper about Satsurblia and Kotias:
Received 20 Jul 2015 | Accepted 15 Oct 2015 | Published 16 Nov 2015

So they send the paper 4 days after paper with CHG was published and of course they didn't have enough time to revise it. Even though they made some revisions in December I guess the amount of rewriting required to include new data was too much.

Anabasis
01-02-2016, 10:55 AM
This paper shows that CHG spread over Anatolia and Eastern Europe during Chalcolithic period. Metal working is probably discovered by CHG like people. Kumtepe people seems %30-35 CHG. And this people seems different from Yamna Like people.

ArmandoR1b
01-02-2016, 11:11 AM
Looks quite good. Kum6 has much CHG-like admixture and situates actually close to Cyprus.


Sorry Chad,
I disagree--Look at K=12. I don't have the blinders on about Yamnaya which interests me little. One can readily see the CHG shift in Kum6 at K=12.


This paper shows that CHG spread over Anatolia and Eastern Europe during Chalcolithic period. Metal working is probably discovered by CHG like people. Kumtepe people seems %30-35 CHG. And this people seems different from Yamna Like people.


There is no comparison with Kotias/Satsurbia? So this paper is written before CHG was known ?

Kumtepe do not have CHG they have West Asian.

Since Kotias/Satsurbia aren't included in the Admixture graph then all we have to compare to is Yamnaya and Late Neolithic which we know from Jones et al. had CHG. At k=12 kum6=Blue + Dark Green (West Asian) + Orange (Neolithic Farmer) + Purple but Yamnaya=Blue + Dark Blue (WHG) + Light Blue or Sky Blue (South Asian) + Teal + Red. The only color shared is the Blue and the only pop that has that as a main component is Tubalar.

In Haak et al. (which has 63 more populations and needs a higher k) k=15 the CHG was Bright Green which is South Asian but at k=16 the Teal showed up in Yamnaya and South Asians but bright Green disappeared from Yamnaya and Europeans but remained at a reduce rate in South Asians.

If both papers would have included Kotias/Satsurbia then we would see yet a different color for CHG. I have no idea what South Asian would show with Kotias/Satsurbia included.

It's also unfortunate that Jones et al. does not include a legible ADMIXTURE graph of the populations that were compared.

Admixture graphs are very useful as long as the right populations are used and they are interpreted correctly. Apart from Kotias/Satsurbia missing this paper doesn't include the Native American and African populations that Haak et al. included so it makes it even harder to compare the two.

Anabasis
01-02-2016, 12:07 PM
Kumtepe do not have CHG they have West Asian...

.

Well i am sorry but i didnt read rest of your post after reading your that statement. Could you tell me the difference between "West Asian" and "CHG"? (Of course there is some but its a fact that we are facing with spread of West Asian/CHG like people in this paper). Both Cypriots and Sardinians have significant CHG admixture too while they are mostly EEF like. So both you and Chad, whats your purpose on insulting an academic study as "bullshit" just because it doesnt fit your agenda?

ArmandoR1b
01-02-2016, 12:12 PM
Well i am sorry but i didnt read rest of your post after reading your that statement. Could you tell me the difference between "West Asian" and "CHG"? (Of course there is some but its a fact that we are facing with spread of West Asian/CHG like people in this paper). Both Cypriots and Sardinians have significant CHG admixture too while they are mostly EEF like. So both you and Chad, whats your purpose on insulting an academic study as "bullshit" just because it doesnt fit your agenda?

The answer to your first question is in the part of my post that you did not read. There was no insult on my part only an objective view of the data.

Viktor Reznov
01-02-2016, 12:19 PM
I wonder if J will make an appearence. CHG is very strong today in the near east , as is J, and they're obviously associated.
Edit:

* Kum6 carries the H2a3 mitochondrial haplogroup (both females, the other Kum4 was not typed)
Really? :(

Arame
01-02-2016, 12:38 PM
After analysing the data I came to the conclusion that there is no sensation here. Everything is in line what we knew and we were expecting.
IMHO this sample will show some CHG in the Davidski's CHG K10 calculator.
It will be ~ 10%. Or maybe even less than 10%
If we look closely the Early Anatolian Neolithic farmers most of them don't have CHG. But few of them had CHG from 1 - 7 %. So after some 1500 years the level of CHG has slowly increased and we see this sample at late Neolithic 4700BC.
Otzi Iceman's age is 3300 BC. But it is still low in CHG. Some 6%. This mean another 1000-1500 year was needed for the slow diffusion of CHG from Anatolia to Appeninian peninsula.

Off course all this number estimates are based on CHG K10.

Gravetto-Danubian
01-02-2016, 12:38 PM
I wonder if J will make an appearence. CHG is very strong today in the near east , as is J, and they're obviously associated.
Edit:

Really? :(

yes only two samples, and both were females

But yes, it's likely that J spread with CHG. But who know what else did secondarily ?

Arame
01-02-2016, 12:42 PM
It seems that files are already available
http://www.ebi.ac.uk/ena/data/view/PRJEB12155

I am impatient to see how accurate were my guesstimates in the CHG K10 :)

Viktor Reznov
01-02-2016, 12:43 PM
Looks like Chad is right. Although in D-stats they show a bunch of Steppe groups are closer to Kum6 than other EEFs. Armenia_BA is most significantly closer to Kum6 than to other EEFs. Kum6 is from 6700 BP, so not much younger than Stuttgart. There's an about 6,000 year old genome from Greece that's EEF. CHG could and other elements have arrived by 3000 BC. We can't decisively say CHG or anything non-EEF didn't arrive in SE Europe till the Metal ages.

It's cool narrative. A close-knit group colonized most of Europe, and from 5000-3000 BC pretty much everyone in Europe and Turkey was apart of this group. Then starting by at least 3000 BC, people from the Steps and West Asia moved into their territory and ended their reign. And Sardinians are the last almost pure ones let. Without aDNA it would be appear the distinctiveness of Sardinians is just because of isolation, but they represent people that took up most of Europe and Turkey for 1,000s of years.

After Stuttgart, Loschbour, Otzei, etc. it was a mystery how North Europe became non-EEF. Because of lots of DNA from east of Germany we know it's Steppe. But now we need DNA from SE Europe and West Asia dating 6,000-3,000yo to find out who mixed with EEF there. Some were Steppe, but the main source came from West Asia. BTW, Generalissimo I don't think it came in historical times. Italy has the lowest IBD with each other for any European country and has lowest IBD with other Europeans. If West Asian or Balkan admixture came in historical times Italians should have more IBD with each other and or West Asia/Balkans. I think it came in sometime between 0 AD and 3000 BC.

You're saying the high CHG to EHG ratio in Europe is not from CHG but rather from a population that was related but different? Because it really does'nt make sense to say it all comes from Yamnaya since Yamnaya had pretty equal amounts of EHG and CHG.

Anabasis
01-02-2016, 12:47 PM
After analysing the data I came to the conclusion that there is no sensation here. Everything is in line what we knew and we were expecting.
IMHO this sample will show some CHG in the Davidski's CHG K10 calculator.
It will be ~ 10%. Or maybe even less than 10%
If we look closely the Early Anatolian Neolithic farmers most of them don't have CHG. But few of them had CHG from 1 - 7 %. So after some 1500 years the level of CHG has slowly increased and we see this sample at late Neolithic 4700BC.
Otzi Iceman's age is 3300 BC. But it is still low in CHG. Some 6%. This mean another 1000-1500 year was needed for the slow diffusion of CHG from Anatolia to Appeninian peninsula.

Off course all this number estimates are based on CHG K10.

I agree with you but i expect around %20-30 CHG. I might be wrong but i am sure that it will have more CHG then Barcin ones.

nuadha
01-02-2016, 12:49 PM
Interesting. Does this add weight to the belief of some that the non-EHG ancestors of Yamnaya peoples were from Central Asia and not from the Caucasus?

I don't see why that would be.

One explanation for the results is that green was true middle eastern neolithic (spreading from iraq to the northwest, north, and east), orange derives from mesolithic ageans, while the non EHG in the yamnaya was still from the caucasus.

I would call this a decent explanation IF the components mentioned could actually be trusted as reflecting ancient populations. Davidski doubts that...

ArmandoR1b
01-02-2016, 01:00 PM
After analysing the data I came to the conclusion that there is no sensation here. Everything is in line what we knew and we were expecting.
IMHO this sample will show some CHG in the Davidski's CHG K10 calculator.
It will be ~ 10%. Or maybe even less than 10%
If we look closely the Early Anatolian Neolithic farmers most of them don't have CHG. But few of them had CHG from 1 - 7 %. So after some 1500 years the level of CHG has slowly increased and we see this sample at late Neolithic 4700BC.
Otzi Iceman's age is 3300 BC. But it is still low in CHG. Some 6%. This mean another 1000-1500 year was needed for the slow diffusion of CHG from Anatolia to Appeninian peninsula.

Off course all this number estimates are based on CHG K10.

It looks like you are confusing the dark green for CHG.

nuadha
01-02-2016, 01:03 PM
I don't see why that would be.

One explanation for the results is that green was true middle eastern neolithic (spreading from iraq to the northwest, north, and east), orange derives from mesolithic ageans, while the non EHG in the yamnaya was still from the caucasus.

I would call this a decent explanation IF the components mentioned could actually be trusted as reflecting ancient populations. Davidski doubts that...

On second thought this admixture map looks pretty much the same as the admixture maps that we used to see 5/6 years ago based on just iceman and swedish hg and a whole bunch of modern populations. I think the admixture results are being dictated by moderns (i.e. drift and other events that happened after kumptepe)

parastais
01-02-2016, 01:08 PM
Dark blue seems to be some component that includes much more than just pure "standard" WHG.
Those dark blue ratios in modern folk seem way inflated for it to be WHG.

Shaikorth
01-02-2016, 01:14 PM
Dark blue seems to be some component that includes much more than just pure "standard" WHG.
Those dark blue ratios in modern folk seem way inflated for it to be WHG.

It's the Baltic component, also often labelled N-European in Gedmatch, which ADMIXTURE often creates based on modern populations. This ofc means it is not WHG but has a Corded Ware and BA steppe shift.

Arame
01-02-2016, 01:25 PM
ArmandoR1b

I don't think dark green is equal to CHG. Dark green is ~35% in Kum6.
CHG will be something =<10% in Davidski's CHG calculator. IMHO.

Arbogan
01-02-2016, 02:43 PM
Basically we can infer from this paper that CHG started appearing in North-West Anatolia during the late stage of the neolithic. That leaves the possibility of them having, appeared in eastern anatolia perhaps even earlier. Basically the expansion of CHG coincides with the establishment of West Asian post-neolithic civilization. Now we just need some early and late neolithic samples from the levant and Iran to confirm this.

The dark green. Can't be anything other than CHG or at least CHG influenced. Based on the diffusion of ancestral groups modern population. I also do not agree that it's Pre-Whg neolithic ancestry either as someone suggested . It would show almost no statistical difference and would overlap too much with the bacon samples.

Edit: looked at the blurred out graphic. The dark green coincides well with the distributionof CHG. It's found in North Western south Asia and south central asia.

vettor
01-02-2016, 06:39 PM
why where all the early Neolithic German farmers excluded in this paper?..........I thought they would be helpful to link them to the anatolians.

Anabasis
01-03-2016, 09:08 AM
Its so clear in paper that Kum6 clustered between Sardinians and Cypriots. Its something like a proof of CHG effect on kum6 which makes Kum6 pull toward Cypriots rather then Sardians. as we know most of the EN aDNAs of Anatolia and Europe were like Sardinian like but in Kum6 we can see that its more like Cypriots. Modern Cypriots have more CHG then Sardianins which make that similarity with Kum6.

One more point about the paper that Yamnaya seems not have CHG like ancestry which is described as Dark Green component in paper. Thus the theories related with southern caucasian effect on yamnaya might be collapsed. Yamnaya seems have Gedrosian type "CHG" which is much more absent in Souther caucasian and mostly found in South central Asians. That might powered the ideas which says roots of the PIEs are somewhere from north India-Gedrosia.

On the other hand i noticed that Kalash have an Teal coloured component which not any other population have.

PS: I wrote every combination of sentences ith my restricted English to make myself more clearer . Many times i closed my eyes and just typed to express my ideas here. Thats why i usualy try to set up simple sentences to be more cleared. But that doesnt mean that i have not deeper thought about the discussions.

Gravetto-Danubian
01-03-2016, 10:27 AM
Its so clear in paper that Kum6 clustered between Sardinians and Cypriots. Its something like a proof of CHG effect on kum6 which makes Kum6 pull toward Cypriots rather then Sardians. as we know most of the EN aDNAs of Anatolia and Europe were like Sardinian like but in Kum6 we can see that its more like Cypriots. Modern Cypriots have more CHG then Sardianins which make that similarity with Kum6.

One more point about the paper that Yamnaya seems not have CHG like ancestry which is described as Dark Green component in paper. Thus the theories related with southern caucasian effect on yamnaya might be collapsed. Yamnaya seems have Gedrosian type "CHG" which is much more absent in Souther caucasian and mostly found in South central Asians. That might powered the ideas which says roots of the PIEs are somewhere from north India-Gedrosia.

On the other hand i noticed that Kalash have an Teal coloured component which not any other population have.

PS: I wrote every combination of sentences ith my restricted English to make myself more clearer . Many times i closed my eyes and just typed to express my ideas here. Thats why i usualy try to set up simple sentences to be more cleared. But that doesnt mean that i have not deeper thought about the discussions.


Yes you make sense ;)
But I think some of the earlier formal stats showed that Yamnaya has "Caucasus" like CHG, and the rest of Europe - esp southern - has South Asian type .
But who knows. 2016 should bring more aDNA from south of the Caucasus, and define everything better

Gravetto-Danubian
01-03-2016, 10:28 AM
Its so clear in paper that Kum6 clustered between Sardinians and Cypriots. Its something like a proof of CHG effect on kum6 which makes Kum6 pull toward Cypriots rather then Sardians. as we know most of the EN aDNAs of Anatolia and Europe were like Sardinian like but in Kum6 we can see that its more like Cypriots. Modern Cypriots have more CHG then Sardianins which make that similarity with Kum6.

One more point about the paper that Yamnaya seems not have CHG like ancestry which is described as Dark Green component in paper. Thus the theories related with southern caucasian effect on yamnaya might be collapsed. Yamnaya seems have Gedrosian type "CHG" which is much more absent in Souther caucasian and mostly found in South central Asians. That might powered the ideas which says roots of the PIEs are somewhere from north India-Gedrosia.

On the other hand i noticed that Kalash have an Teal coloured component which not any other population have.

PS: I wrote every combination of sentences ith my restricted English to make myself more clearer . Many times i closed my eyes and just typed to express my ideas here. Thats why i usualy try to set up simple sentences to be more cleared. But that doesnt mean that i have not deeper thought about the discussions.


Yes you make sense ;)
But I think some of the earlier formal stats showed that Yamnaya has "Caucasus" like CHG, and the rest of Europe - esp southern - has South Asian type .
But who knows. 2016 should bring more aDNA from south of the Caucasus, and define everything better

Anabasis
01-03-2016, 10:48 AM
Yes you make sense ;)
But I think some of the earlier formal stats showed that Yamnaya has "Caucasus" like CHG, and the rest of Europe - esp southern - has South Asian type .
But who knows. 2016 should bring more aDNA from south of the Caucasus, and define everything better

Do you have any source for your statement that southern europeans have "South Asian "type CHG ancestry?

Generalissimo
01-03-2016, 11:17 AM
There's no South Asian ancestry in Europe, unless you mean Roma. Not ANI, and nothing via the steppe.

Links between the European steppe and South Asia date to the Middle to Late Bronze Age, when there were massive population movements from the steppe to South Central Asia, resulting in founder effects which now create these funny Admixture clusters.

There are strong IBD and Y-chomosome links between the MB/LB steppe and South Asia, but the traffic is all one way to the south.

Gravetto-Danubian
01-03-2016, 11:31 AM
I meant to say south Central Asia, whose links with Europe from the steppe and via Anatolia.

bored
01-03-2016, 12:52 PM
I meant to say south Central Asia, whose links with Europe from the steppe and via Anatolia.

South Central Asia isn't a completely separate entity from South Asia. I don't think there is South Central Asia specific CHG that's not found in South Asians, for example.

rozenfeld
01-03-2016, 12:56 PM
Which region are you calling South Central Asia? Uzbekistan + Tajikistan?

Gravetto-Danubian
01-03-2016, 12:56 PM
South Central Asia isn't a completely separate entity from South Asia. I don't think there is South Central Asia specific CHG that's not found in South Asians, for example.

Today, yes. However I was using SCA as a specific referential point for the said prehistoric processes

Awale
01-03-2016, 01:11 PM
Which region are you calling South Central Asia? Uzbekistan + Tajikistan?

I tend to understand it as Pakistan+Afghanistan (http://2.bp.blogspot.com/-0EDuUFy6IWE/Vl7woUdVLiI/AAAAAAAADtk/acgpW5jgrnI/s1600/South-Central%2BAsia.png) (with perhaps Tajikistan), everything else is "East-Central Asia", I suppose. Honestly, these aren't "real regions" even used by anyone "in real-life" (as far as I know...) but something we anthro-genomics enthusiasts cooked up cos it makes sense with some of the genetics.

jesus
01-03-2016, 01:18 PM
South Central Asia is basically the non Turkic, Iranic speaking parts of Central Asia. Or the Eastern/North Eastern parts of the Iranian plateau. Big parts of South Central Asia were/still known as Khorasan.

bored
01-03-2016, 01:27 PM
I tend to understand it as Pakistan+Afghanistan (http://2.bp.blogspot.com/-0EDuUFy6IWE/Vl7woUdVLiI/AAAAAAAADtk/acgpW5jgrnI/s1600/South-Central%2BAsia.png) (with perhaps Tajikistan), everything else is "East-Central Asia", I suppose. Honestly, these aren't "real regions" even used by anyone "in real-life" (as far as I know...) but something we anthro-genomics enthusiasts cooked up cos it makes sense with some of the genetics.

Not all of Pakistan. Only Iranic and northern Dardic areas. Neighbouring Punjab is an integral part of South Asia.

jesus
01-03-2016, 01:29 PM
I tend to understand it as Pakistan+Afghanistan (http://2.bp.blogspot.com/-0EDuUFy6IWE/Vl7woUdVLiI/AAAAAAAADtk/acgpW5jgrnI/s1600/South-Central%2BAsia.png) (with perhaps Tajikistan), everything else is "East-Central Asia", I suppose. Honestly, these aren't "real regions" even used by anyone "in real-life" (as far as I know...) but something we anthro-genomics enthusiasts cooked up cos it makes sense with some of the genetics.

Agreed, it's not a popular term in real life conversions, however the term Khurasan is widely used by Afghans/Tajiks/Eastern Iranians.

Core Khorasan

http://www.khorasanzameen.net/images/louvre/khurasan1.jpg


Greater Khorasan


http://s5.picofile.com/file/8124264142/KhorasanMap_Large.jpg

http://s21.photobucket.com/user/Paniranism/media/KhorasanMap.jpg.html

ArmandoR1b
01-03-2016, 02:58 PM
The study says exactly what I have been saying about the ADMIXTURE results which is that the green is West Asian and not what is found in Yamnaya which is light blue. The light blue is CHG. See pages 3 and 4.


The first one (blue), observed
as the main component in all hunter-gatherers, is also found
as a minor contribution to all farmers, which is in line with
the observed admixture from hunter-gatherers into farmers [3].
The second (orange) and the third (green) components were
observed mostly in farmers to varying degrees (5%–68% and
0.06%–45% for K = 9, respectively). The orange component
is mainly found in present-day Western Europeans, whereas
the third component (green) is mostly found in the modern-day
Near East and Caucasus, and the highest proportion of this third
component among Neolithic individuals was observed in Kum6
(45% for K = 9). The notion that this component is West Asian
is also supported by its presence in a Bronze Age Armenian
sample (51%), which contains less than 2% of the orange
component. Interestingly, this ‘‘West Asian’’ component (green)
is not related to the potential genetic material brought to Europe
by migration during the Bronze Age and recently connected to
the Yamnaya culture [19, 24], visualized in Figure 3 as light
blue, and it is observed in high frequency in modern-day people
from southern Asia. The elevated ‘‘West Asian’’ affinity of Kum6
is likely to be the cause of the genetic differentiation observed
between Kum6 and all other ancient farmers shown in the PCA
plot (Figure 1). A clear decline was observed in the values of
the green component over time (average of 29% in Early
Neolithic, 14% in Middle Neolithic, and 2% in Late Neolithic),
which is consistent with increased admixture with hunter-gath-
erer groups [3]. Our results suggest that
the two ancestry components of ancient
farmers (orange and green in Figure 3)
were established at an early stage, probably
before the first farmers expanded
into Europe, and were maintained in Europe
up until the end of Middle Neolithic
and that both components are present
in various modern-day European populations.
Therefore, these observations
directly link the early European Neolithic
gene pool to western Anatolia.

This means that it is not a CHG effect on kum6 which makes Kum6 pull toward Cypriots rather then Sardinians.

Anabasis
01-03-2016, 04:06 PM
The study says exactly what I have been saying about the ADMIXTURE results which is that the green is West Asian and not what is found in Yamnaya which is light blue. The light blue is CHG. See pages 3 and 4.



This means that it is not a CHG effect on kum6 which makes Kum6 pull toward Cypriots rather then Sardinians.

Because of the passage you quoted, Kum6 have CHG effect, because West Asian= CHG. Caucaus Hunter Gatherers discovered in S. caucasia and that component is found along modern caucasians in high rates. (in SCA as well) Thus with no doubt "This means that it is a CHG effect on kum6 which makes Kum6 pull toward Cypriots rather then Sardinians"

Dr_McNinja
01-03-2016, 04:18 PM
There's no South Asian ancestry in Europe, unless you mean Roma. Not ANI, and nothing via the steppe.

Links between the European steppe and South Asia date to the Middle to Late Bronze Age, when there were massive population movements from the steppe to South Central Asia, resulting in founder effects which now create these funny Admixture clusters.

There are strong IBD and Y-chomosome links between the MB/LB steppe and South Asia, but the traffic is all one way to the south.
The little bit of "ASE" that populations like EHG and Yamnaya have, what route did it come from? Surely that went into Europe, isn't that why Europeans score minor amounts of South Asian or Gedrosian?

kingjohn
01-03-2016, 04:32 PM
when i put the yamnaya gedmatch kits in eurogenes k36 tool {which is great }
they score 20% north caucasus and 10-12 % south central asia .
but when i put kotias in eurogenes k36 he have a 46% north caucasus and 39% west caucasus and 12% south central asia.
all the yamnaya samples carry north caucasus and south central asia they dont carry west caucaus.
it means that in yamnaya the non european chg is expressed {by those 2 components only}
the south central asia component is the real link between pakistan afghanistan and the yamnaya from the north.
regards
adam

Arbogan
01-03-2016, 04:38 PM
I think we need to put the kumtepe into admixture or D-statistis. I think it's just going to turn out to be a ANF modified by CHG. The neolithic Iranian samples will confirm the same as well.

Anabasis
01-03-2016, 05:05 PM
I think we need to put the kumtepe into admixture or D-statistis. I think it's just going to turn out to be a ANF modified by CHG. The neolithic Iranian samples will confirm the same as well.

Anyone upload Kum6 into gedmatch?

ArmandoR1b
01-03-2016, 05:25 PM
Because of the passage you quoted, Kum6 have CHG effect, because West Asian= CHG. Caucaus Hunter Gatherers discovered in S. caucasia and that component is found along modern caucasians in high rates. (in SCA as well) Thus with no doubt "This means that it is a CHG effect on kum6 which makes Kum6 pull toward Cypriots rather then Sardinians"

What I have been pointing out is that West Asian does not equal CHG, The dark green is West Asian and the light blue is CHG just as the authors of the study have pointed out.

Gravetto-Danubian
01-04-2016, 12:06 AM
Which region are you calling South Central Asia? Uzbekistan + Tajikistan?

In my mind I have the broad region of northern Iran, the mountainous littoral of the -Stans and adjacent Hindu Kush.

As Awale rightly mentioned, these are heuristic, non-specific terms, and Im fully aware the Indus basin is actually considered 'South Asia'.
And again, its not reflective of any modern cultural or political divisioning, but one can appreciate how these regions might have shared pertinence in Bronze Age dynamics.

Generalissimo
01-04-2016, 02:36 AM
All the yamnaya samples carry north caucasus and south central asia they dont carry west caucaus.
it means that in yamnaya the non european chg is expressed {by those 2 components only}
the south central asia component is the real link between pakistan afghanistan and the yamnaya from the north.

ADMIXTURE is not a formal mixture test, but despite that many people here use it like that, as per above, which is a mistake that will lead you to wrong conclusions.

Kurd
01-04-2016, 03:43 AM
The little bit of "ASE" that populations like EHG and Yamnaya have, what route did it come from? Surely that went into Europe, isn't that why Europeans score minor amounts of South Asian or Gedrosian?

As has been stated here, the overwhelming amount of evidence indicates that the geneflow between the steppe and S Asia was a one way flow from N to S.

If EHG or Yamanya does show a little ASE in some calculators, there could be 2 possible explanations:

1- It could be due to MA1, as EHG is substantially MA1 derived to the best of my knowledge. The ASE signal in MA1 is very old, and it is probably from a source basal to MA1 and S Asians, perhaps a paleolithic ANE source.

2- ADMIXTURE is not informative for direction of geneflow. It may be picking up the EHG alleles conferred to S Asia via steppe groups, for whom EHG was ancestral.

In either case, this should not be seen as post paleolithic geneflow from S Asia to Europe. Personally, I have never seen EHG or Yamnaya score any ASI or SE Asian in any of my calculators. My calculators have shown MA1 or Ust Ishim score ASI, but that is likely due to the reason mentioned under (1). Also, for confirmation of any admixture one should defer to formal tests, as has also been stated here.

EDIT: I have not seen Europeans score any ASI, beyond what can be considered noise. Gedrosian is a different strory. Prior to CHG some calculators, including mine, had separate Caucausus and Gedrosian (sometimes referred to as ANI) components. The Gedrosian component was based on caucausus derived alleles in the drifted Baloch-Brahui-Makrani populations. The gedrosian and caucausus cluster are separated by small GD (I remember it was around 0.04). This overlap between the 2 clusters is what I believe causes Europeans to score small amounts of Gedrosian, which in fact is Caucausus related admixture.

parasar
01-04-2016, 03:59 AM
There's no South Asian ancestry in Europe, unless you mean Roma. Not ANI, and nothing via the steppe.

Links between the European steppe and South Asia date to the Middle to Late Bronze Age, when there were massive population movements from the steppe to South Central Asia, resulting in founder effects which now create these funny Admixture clusters.

There are strong IBD and Y-chomosome links between the MB/LB steppe and South Asia, but the traffic is all one way to the south.

"Novosvobodnaja fanden wir auch die Haplogruppe М52"
https://www.academia.edu/16867978/2015_Der_Kaukasus_im_Spannungsfeld_zwischen_Osteur opa_und_Vorderem_Orient_Dialog_der_Kulturen_Kultur _des_Dialoges_im_Gedenken_an_den_140._Geburtstag_v on_Alexander_A._Miller_Materialien_der_Internation alen_Fachtagung_Humboldt-Kolleg_In_Germany_English_and_Russian_

vettor
01-04-2016, 05:19 AM
There's no South Asian ancestry in Europe, unless you mean Roma. Not ANI, and nothing via the steppe.

Links between the European steppe and South Asia date to the Middle to Late Bronze Age, when there were massive population movements from the steppe to South Central Asia, resulting in founder effects which now create these funny Admixture clusters.

There are strong IBD and Y-chomosome links between the MB/LB steppe and South Asia, but the traffic is all one way to the south.

There is L-M20 in the south alps, estonia, portugal and netherlands which is South-asia
there is H2-p96 in central europe which is south-asia

you need to look at all of Europe and not just concentrate on eastern europe and the steppe

Anabasis
01-04-2016, 05:46 AM
There is L-M20 in the south alps, estonia, portugal and netherlands which is South-asia
there is H2-p96 in central europe which is south-asia

you need to look at all of Europe and not just concentrate on eastern europe and the steppe

The L M20 in baltic is L2 and its too old. Probably mesoloic clade which is not found in south asia. On the other hand Lm317 in italy is west asian clade which seems neolethic or early bronze age clade of L in europe. its too rare in south asia too.

jesus
01-04-2016, 06:07 AM
The L M20 in baltic is L2 and its too old. Probably mesoloic clade which is not found in south asia. On the other hand Lm317 in italy is west asian clade which seems neolethic or early bronze age clade of L in europe. its too rare in south asia too.

Are you sure Etonian L is L2 ? as far as I know there are no Estonian samples in the L project. https://www.familytreedna.com/public/Y-Haplogroup-L/default.aspx?section=yresults

Italy has all three major subclades of L it seems, but most are under L2. There is no L in mesolithic and neolithic Europe/west Asia, it's probably fairly recent in Europe. There is only one L2 sample in Indo Aryan and Dravidian speakers from south Asia, that sample is from Punjab, Pakistan.

Edit: found some info about the Estonian L




L-L595 is found only in Europe from Ireland, Iberian Peninsula and Sardinia to Eastern Europe and their highest frequency is found in Estonia,[Footnote 5] Scozzari 2001, Lappalainen 2008.





https://en.wikipedia.org/wiki/Haplogroup_L-M20#L-M295

Anabasis
01-04-2016, 06:36 AM
Are you sure Etonian L is L2 ? as far as I know there are no Estonian samples in the L project. https://www.familytreedna.com/public/Y-Haplogroup-L/default.aspx?section=yresults

Italy has all three major subclades of L it seems, but most are under L2. There is no L in mesolithic and neolithic Europe/west Asia, it's probably fairly recent in Europe. There is only one L2 sample in Indo Aryan and Dravidian speakers from south Asia, that sample is from Punjab, Pakistan.

Edit: found some info about the Estonian L




https://en.wikipedia.org/wiki/Haplogroup_L-M20#L-M295

Lm20 s in italy are mostly L1b m317 and mostly M349 under M317. Those clades are west asian clades. Found also cyprus and lubanon. in baltic countries its L2 or L 595 which is very far from all known L clades in south Asia. We need full seuence test to determain the age. İ guess L595 are more then 20k years old.

Ryukendo
01-04-2016, 08:05 AM
Well i am sorry but i didnt read rest of your post after reading your that statement. Could you tell me the difference between "West Asian" and "CHG"? (Of course there is some but its a fact that we are facing with spread of West Asian/CHG like people in this paper). Both Cypriots and Sardinians have significant CHG admixture too while they are mostly EEF like. So both you and Chad, whats your purpose on insulting an academic study as "bullshit" just because it doesnt fit your agenda?


Its so clear in paper that Kum6 clustered between Sardinians and Cypriots. Its something like a proof of CHG effect on kum6 which makes Kum6 pull toward Cypriots rather then Sardians. as we know most of the EN aDNAs of Anatolia and Europe were like Sardinian like but in Kum6 we can see that its more like Cypriots. Modern Cypriots have more CHG then Sardianins which make that similarity with Kum6.

One more point about the paper that Yamnaya seems not have CHG like ancestry which is described as Dark Green component in paper. Thus the theories related with southern caucasian effect on yamnaya might be collapsed. Yamnaya seems have Gedrosian type "CHG" which is much more absent in Souther caucasian and mostly found in South central Asians. That might powered the ideas which says roots of the PIEs are somewhere from north India-Gedrosia.

On the other hand i noticed that Kalash have an Teal coloured component which not any other population have.

PS: I wrote every combination of sentences ith my restricted English to make myself more clearer . Many times i closed my eyes and just typed to express my ideas here. Thats why i usualy try to set up simple sentences to be more cleared. But that doesnt mean that i have not deeper thought about the discussions.


Anabiasis, it is extremely important not to confuse the inferences of ADMIXTURE with actual ancestral populations. ADMIXTURE doesn't usually come up with a 100% WHG component, 100% ENF component etc, unless we add ancient samples to the dataset; and even then it generally has difficulty representing the ancestral populations exactly. It will be an incredible wonder if this ADMIXTURE run produces a 'CHG' component without a CHG genome even being included.

Starting from K=8 when the orange and green components split, we can tell that the 'Dark Green' component in the ADMIXTURE is in fact not rich in CHG ancestry, because populations such as Sindhi, Yamnaya, Kalash, or Estonian, which should have high CHG, have ~0% of the Dark green component, and others like Tajik Pomiri and Baloch have far too low levels of it to represent their CHG level accurately. These populations instead have high 'N European' and 'SC Asian' components. Supporting this inference, the N European and SC Asian components appear in the Caucasus, Greece and Anatolia at far higher levels than are usually seen. North Ossetes are almost 70% N Euro+SC Asian combined. So it seems that 'N European' and 'SC Asian' swallow up CHG allele frequencies, which is consistent with Lithuanians and Punjabi scoring ~100% in N European and SC Asian respectively, despite having complex ancestry with CHG contributions.

Instead, the Dark Green component represents a Mediterranean affinity, like the orange component, just a different one from the 'EEF' Sardinian-centered one we're used to seeing. For example, Balochi and Brahui are generally more 'Mediterranean' than other populations in S and C Asia, and they are differentiated from the rest of the S and C Asians, in fact, by carrying high levels of the the Dark Green component, which other SC Asians do not...

To support this assertion further, look at the behaviour of the Abkhaz and Adygei as K increases from 5 to 12. The 'SC Asian' component gradually retreats from these populations at each increase in K, and the Dark green component increases, becoming more and more CHG-rich. As it does so, the dark green component steadily increases in the Tajik Pomiri, as is expected, but it retreats from neolithic genomes, including the Kumtepe, from K=5 to 12, indicating that, as the similarity to CHG increases, the similarity to the neolithic genomes decreases...

That said, a small amount of Dark Green is left in the KumTepe after it disappears from all the other neolithic genomes, so I am open to there being a small amount of CHG in KumTepe; but the formal statistics show that this isn't the significant factor influencing KumTepe's similarity to LNBAs here, as it is the EEF-rich LNBAs that show the similarity to KumTepe, not the CHG-rich ones.

Ryukendo
01-04-2016, 08:08 AM
Yes you make sense ;)
But I think some of the earlier formal stats showed that Yamnaya has "Caucasus" like CHG, and the rest of Europe - esp southern - has South Asian type .
But who knows. 2016 should bring more aDNA from south of the Caucasus, and define everything better

GD,

Where did you see this?

Gravetto-Danubian
01-04-2016, 09:48 AM
GB,

Where did you see this?

Hi Ryu

Naturally, I (thought) i read it somewhere along the conversations amongst yourself and the other fellows on Eurogenes. I can't find the eaxct thread.
Maybe i misinterpreted it, or got things vice-versa, but

* maybe there are subcomponents within this 'CHG'

* I am willing to bet that there were two, possibly 3, routes of entry into Europe of CHG .

ArmandoR1b
01-04-2016, 04:00 PM
Anabiasis, it is extremely important not to confuse the inferences of ADMIXTURE with actual ancestral populations. ADMIXTURE doesn't usually come up with a 100% WHG component, 100% ENF component etc, unless we add ancient samples to the dataset; and even then it generally has difficulty representing the ancestral populations exactly. It will be an incredible wonder if this ADMIXTURE run produces a 'CHG' component without a CHG genome even being included.

Starting from K=8 when the orange and green components split, we can tell that the 'Dark Green' component in the ADMIXTURE is in fact not rich in CHG ancestry, because populations such as Sindhi, Yamnaya, Kalash, or Estonian, which should have high CHG, have ~0% of the Dark green component, and others like Tajik Pomiri and Baloch have far too low levels of it to represent their CHG level accurately. These populations instead have high 'N European' and 'SC Asian' components. Supporting this inference, the N European and SC Asian components appear in the Caucasus, Greece and Anatolia at far higher levels than are usually seen. North Ossetes are almost 70% N Euro+SC Asian combined. So it seems that 'N European' and 'SC Asian' swallow up CHG allele frequencies, which is consistent with Lithuanians and Punjabi scoring ~100% in N European and SC Asian respectively, despite having complex ancestry with CHG contributions.

Instead, the Dark Green component represents a Mediterranean affinity, like the orange component, just a different one from the 'EEF' Sardinian-centered one we're used to seeing. For example, Balochi and Brahui are generally more 'Mediterranean' than other populations in S and C Asia, and they are differentiated from the rest of the S and C Asians, in fact, by carrying high levels of the the Dark Green component, which other SC Asians do not...

To support this assertion further, look at the behaviour of the Abkhaz and Adygei as K increases from 5 to 12. The 'SC Asian' component gradually retreats from these populations at each increase in K, and the Dark green component increases, becoming more and more CHG-rich. As it does so, the dark green component steadily increases in the Tajik Pomiri, as is expected, but it retreats from neolithic genomes, including the Kumtepe, from K=5 to 12, indicating that, as the similarity to CHG increases, the similarity to the neolithic genomes decreases...

That said, a small amount of Dark Green is left in the KumTepe after it disappears from all the other neolithic genomes, so I am open to there being a small amount of CHG in KumTepe; but the formal statistics show that this isn't the significant factor influencing KumTepe's similarity to LNBAs here, as it is the EEF-rich LNBAs that show the similarity to KumTepe, not the CHG-rich ones.

The above is an example of ADMIXTURE results being correctly analyzed. I had pretty much stated that same in much less words and details and was also the same as what the authors had stated in the study. The Dark Green is a West Asian component that forms part of the Neolithic and not a part of Yamnaya. If there is any CHG in the European Neolithic it is very minor.

Anabasis
01-04-2016, 04:25 PM
The above is an example of ADMIXTURE results being correctly analyzed. I had pretty much stated that same in much less words and details and was also the same as what the authors had stated in the study. The Dark Green is a West Asian component that forms part of the Neolithic and not a part of Yamnaya. If there is any CHG in the European Neolithic it is very minor.

Well problem is in your assumption that the Gedrosian type admixture in Yamnaya is CHG but rest is not. But i insist that we can not nammed it as CHG as far as same component also found in South Caucasian population. Thus we should split CHG into two part. One is South caucasian and the other one is Gedrosia. As far as you dont split it into two different component you can not exclude one as "west asian" but not CHG. Thats my my point.

Dr_McNinja
01-04-2016, 05:08 PM
As has been stated here, the overwhelming amount of evidence indicates that the geneflow between the steppe and S Asia was a one way flow from N to S.

If EHG or Yamanya does show a little ASE in some calculators, there could be 2 possible explanations:

1- It could be due to MA1, as EHG is substantially MA1 derived to the best of my knowledge. The ASE signal in MA1 is very old, and it is probably from a source basal to MA1 and S Asians, perhaps a paleolithic ANE source.

2- ADMIXTURE is not informative for direction of geneflow. It may be picking up the EHG alleles conferred to S Asia via steppe groups, for whom EHG was ancestral.

In either case, this should not be seen as post paleolithic geneflow from S Asia to Europe. Personally, I have never seen EHG or Yamnaya score any ASI or SE Asian in any of my calculators. My calculators have shown MA1 or Ust Ishim score ASI, but that is likely due to the reason mentioned under (1). Also, for confirmation of any admixture one should defer to formal tests, as has also been stated here.

EDIT: I have not seen Europeans score any ASI, beyond what can be considered noise. Gedrosian is a different strory. Prior to CHG some calculators, including mine, had separate Caucausus and Gedrosian (sometimes referred to as ANI) components. The Gedrosian component was based on caucausus derived alleles in the drifted Baloch-Brahui-Makrani populations. The gedrosian and caucausus cluster are separated by small GD (I remember it was around 0.04). This overlap between the 2 clusters is what I believe causes Europeans to score small amounts of Gedrosian, which in fact is Caucausus related admixture.#1 is a possibility. It would also explain why it's that small amount.

However, the Bronze Age Central Asian groups like Andronovo, Afansievo, Sintashta, etc have nearly double that amount of ASE (8-11%) which they presumably picked up in Central Asia. Which would imply there were even greater degrees of that admixture there. Or that it was connected to high ANE, which they all had as well.

If that's the case, how much of the ASE in South Asia is from ANE? South Asia still has high ANE across the board.

parasar
01-04-2016, 06:41 PM
As has been stated here, the overwhelming amount of evidence indicates that the geneflow between the steppe and S Asia was a one way flow from N to S.

If EHG or Yamanya does show a little ASE in some calculators, there could be 2 possible explanations:

1- It could be due to MA1, as EHG is substantially MA1 derived to the best of my knowledge. The ASE signal in MA1 is very old, and it is probably from a source basal to MA1 and S Asians, perhaps a paleolithic ANE source.

2- ADMIXTURE is not informative for direction of geneflow. It may be picking up the EHG alleles conferred to S Asia via steppe groups, for whom EHG was ancestral.

In either case, this should not be seen as post paleolithic geneflow from S Asia to Europe. Personally, I have never seen EHG or Yamnaya score any ASI or SE Asian in any of my calculators. My calculators have shown MA1 or Ust Ishim score ASI, but that is likely due to the reason mentioned under (1). Also, for confirmation of any admixture one should defer to formal tests, as has also been stated here.

EDIT: I have not seen Europeans score any ASI, beyond what can be considered noise. Gedrosian is a different strory. Prior to CHG some calculators, including mine, had separate Caucausus and Gedrosian (sometimes referred to as ANI) components. The Gedrosian component was based on caucausus derived alleles in the drifted Baloch-Brahui-Makrani populations. The gedrosian and caucausus cluster are separated by small GD (I remember it was around 0.04). This overlap between the 2 clusters is what I believe causes Europeans to score small amounts of Gedrosian, which in fact is Caucausus related admixture.

I agree with 1 and 2 above, but I can't see how MA1 can have any ASI unless ancestral ASI (is it being termed ASE?) is being referred to. But if ancestral ASI (Ust-Ishim or close) is being referred to then MA1=ANE should be nearly 100% derived from ancestral ASI.

ArmandoR1b
01-04-2016, 06:44 PM
Well problem is in your assumption that the Gedrosian type admixture in Yamnaya is CHG but rest is not. But i insist that we can not nammed it as CHG as far as same component also found in South Caucasian population. Thus we should split CHG into two part. One is South caucasian and the other one is Gedrosia. As far as you dont split it into two different component you can not exclude one as "west asian" but not CHG. Thats my my point.

I'm not quite sure why you think the Dark Green can't be separated as West Asian when it is not even found in Yamnaya and it is found in all of the Neolithic people.

In a previous post of yours you confused the Dark Green as being CHG, which it is not, and it seems you are still confusing Dark Green as CHG and is the cause of this discussion to be on going.


One more point about the paper that Yamnaya seems not have CHG like ancestry which is described as Dark Green component in paper. Thus the theories related with southern caucasian effect on yamnaya might be collapsed.

The paper does NOT describe Dark Green as CHG. The paper states;


the third component (green) is mostly found in the modern-day
Near East and Caucasus, and the highest proportion of this third
component among Neolithic individuals was observed in Kum6
(45% for K = 9). The notion that this component is West Asian
is also supported by its presence in a Bronze Age Armenian
sample (51%), which contains less than 2% of the orange
component. Interestingly, this ‘‘West Asian’’ component (green)
is not related to the potential genetic material brought to Europe
by migration during the Bronze Age and recently connected to
the Yamnaya culture [19, 24], visualized in Figure 3 as light
blue

You are unwilling to accept what the authors have stated and you continually state the opposite. It's very clear from the results what the Dark Green is and it has nothing to do with CHG. The Yamnaya do have CHG, contrary to what you have posted in the past.

Kurd
01-04-2016, 07:18 PM
#1 is a possibility. It would also explain why it's that small amount.

However, the Bronze Age Central Asian groups like Andronovo, Afansievo, Sintashta, etc have nearly double that amount of ASE (8-11%) which they presumably picked up in Central Asia. Which would imply there were even greater degrees of that admixture there. Or that it was connected to high ANE, which they all had as well.

If that's the case, how much of the ASE in South Asia is from ANE? South Asia still has high ANE across the board.

First, a few clarifications. I noticed that there are a couple of definitions of ANE being used. For example, I noticed in the MDLP calculator has an ANE component, that peaks at around 80% in the Kalash Pathan and Balochi. At the same time those populations score single digit Caucausus. Also, the FST table shows this so called ANE to be closest to ENF and Caucausus components, separated by only 0.04, and 0.024 from those components, respectively. The ANE scores are coincidentally too similar to the Gedrosian scores for those groups The ANE, I am referring to is way too old to have a distance of only 0.024 from modern Caucausus populations, considering the long drift history of those groups since MA1.

My understanding and definition of ANE is something much more ancient. The following table shows the samples I declared ancestral for my supervised Eurasia 9 ASI run:



Population
SE_Asian
ENF
SW_Asian
ASI
EHG
WHG
Siberian_E_Asian
CHG
W_African


Ami
100
0
0
0
0
0
0
0
0


Anatolia_Neolithic
0
100
0
0
0
0
0
0
0


BedouinB
0
0
100
0
0
0
0
0
0


Onge
0
0
0
100
0
0
0
0
0


Karelia_HG
0
0
0
0
100
0
0
0
0


Loschbour
0
0
0
0
0
100
0
0
0


Nganasan
0
0
0
0
0
0
100
0
0


SATSURBILA
0
0
0
0
0
0
0
100
0


Yoruba
0
0
0
0
0
0
0
0
100



The following table shows some run results for a few relevant groups:



Population
SE_Asian
ENF
SW_Asian
ASI
EHG
WHG
Siberian_E_Asian
CHG
W_African


Andronovo (avg)
0.08
12.27
0.14
4.64
41.77
33.03
1.1
6.97
0


Sintashta (avg)
0
13.29
1.71
1.02
37.92
39.7
0
5.05
1.3


MA1
0
0
0
19.54
75.26
0
5.2
0
0


MOTA
1.13
0
8.69
3.79
0
0.59
0.83
3.92
81.05


KOTIAS
0
13.4
5
0
16.04
0
0
62.89
2.67


Great_Andamanese
20.35
0.47
0.9
63.54
1.07
0.7
3.11
9.85
0.01


Paniyas
21.82
0
0
39.19
0
0
0
38.72
0.27


Kharia
38.92
0
0
30.12
0
0
0
29.69
1.27


Naga
73.15
0
0
3.5
0
0
23.35
0
0


Puliyar
15.82
0
0
43.46
0
0
0
40.71
0


Pashtun_Afghan
4.95
8.7
6.65
9.15
11.97
7.32
2.35
48.75
0.16


Punjabi
6.56
3.72
2.58
15.04
10.14
4.35
1.49
55.9
0.23


Tajik_Pomiri
4.73
7.49
5.76
6.08
16.33
14.19
3.17
42.22
0.03


Balochi
1.62
6.97
11.29
9.91
5.94
1.9
1.16
59.83
1.38



The CHG component here includes "Gedrosian" since Gedrosian populations were also used to source its allele frequencies, and also spillover from a ghost basal CHG-S Indian, hitherto undefined population, which I have discussed in other threads. So don't pay too much attention to it. Instead, focus on ASI, SE Asian and EHG. For Andronovo, ASI was 4.64, and Sintashta it was only 1.02. SE Asian was 0 for both. For MA1, ASI was 19.54. I believe that it is more likely that this ASI in Andronovo and Sintashta would have flowed down from MA1, than have been received from India. AS to whether it is actual ASI, the component modal to Onge, or something similar. It is certainly possible that the ASI in MA1, or even perhaps some of the stuff that is showing up as EHG is from a source basal to both ASI and MA1.

The only way to tease it out would be to have a proper ancient sample which is substantially derived from that basal source.

In Admixture, without a proper reference for this basal source, alleles associated with this basal source will defer to the next closest component, in this case ASI

Dr_McNinja
01-04-2016, 07:36 PM
I agree with 1 and 2 above, but I can't see how MA1 can have any ASI unless ancestral ASI (is it being termed ASE?) is being referred to. But if ancestral ASI (Ust-Ishim or close) is being referred to then MA1=ANE should be nearly 100% derived from ancestral ASI.

By ASE I meant the ASE component in Eurogenes K7/K8.

A bunch of these ancient genomes have it, in a pretty ordered gradient. CHG had 5%. The Gedrosian component had 9-10%. Bronze Age Central Asian groups like Afanasievo/Andronovo/Sintashta had it at 8-11%. EHG had it at around 5%. MA1 had 11%.

Modern Caucasus had it at ~4%. Modern Europeans were around 1%.

Actually, in the K7 calculator some of the Afanasievo genomes on Gedmatch (though I think they had low coverage) look like MA1's result with a slight shift towards ENF by 15%.

It could be that it's just what the name implies, "South Eurasian". It doesn't necessarily represent an actual population. It just captures similarity to a layer of admixture, including a certain portion of ASI. ASE hits ~50% in Southeast Asians and ~10-15% in Northeast Asians. And "ASI" as we understand it would be a combination of ASE and Oceanian/Papuan in all likelihood.

But why is there this layer of East Eurasian-like admixture that has this relationship to ANE which is spread all over Asia and captures a lot of ASI? I guess it depends on where MA1 is relative to East Eurasian on the tree.

Btw, you can play around with these base components a bit, for example:

https://docs.google.com/spreadsheets/d/1XGY-UIBC2GcUWGl1FizNvu9ofle4ZPHvfNmbrVYxGMo/edit#gid=1118651176

Having a Gedrosian like component (~10% ASE) but dialing down the ANE just a little results in a component whose spread is like an Iranian version of Gedrosian, with South Asian populations almost always opting for the higher ANE components (which conveniently also describes the more ancient components). Wish I hadn't deleted that calculator.

So I think this reinforces Admixture's non-real nature and the best way to approach South Asia might be to reduce it to its component parts (split its East/West/Southeast parts).

Anabasis
01-04-2016, 08:19 PM
I'm not quite sure why you think the Dark Green can't be separated as West Asian when it is not even found in Yamnaya and it is found in all of the Neolithic people.

In a previous post of yours you confused the Dark Green as being CHG, which it is not, and it seems you are still confusing Dark Green as CHG and is the cause of this discussion to be on going.



The paper does NOT describe Dark Green as CHG. The paper states;



You are unwilling to accept what the authors have stated and you continually state the opposite. It's very clear from the results what the Dark Green is and it has nothing to do with CHG. The Yamnaya do have CHG, contrary to what you have posted in the past.

Oke i got you. Earth turning around and rise on Yamnaya, Birds sing thier songs for it. Every roads in italy links to Yamnaya, Chiniese built great wall for Yamnaya and CHG is nothing without Yamnaya.

Leeroy Jenkins
01-04-2016, 08:29 PM
Oke i got you. Earth turning around and rise on Yamnaya, Birds sing thier songs for it. Every roads in italy links to Yamnaya, Chiniese built great wall for Yamnaya and CHG is nothing without Yamnaya.

No, this really is not about Yamnaya being the focal point of many forum members and overshadowing everything else. Simply put, since there is no green component in Yamnaya and there is green in non-CHG carrying Neolithic European samples, then green in the Kumtepe study's ADMIXTURE is not the "teal" typical of CHG.

parasar
01-04-2016, 08:33 PM
By ASE I meant the ASE component in Eurogenes K7/K8.

A bunch of these ancient genomes have it, in a pretty ordered gradient. CHG had 5%. The Gedrosian component had 9-10%. Bronze Age Central Asian groups like Afanasievo/Andronovo/Sintashta had it at 8-11%. EHG had it at around 5%. MA1 had 11%.

Modern Caucasus had it at ~4%. Modern Europeans were around 1%.

Actually, in the K7 calculator some of the Afanasievo genomes on Gedmatch (though I think they had low coverage) look like MA1's result with a slight shift towards ENF by 15%.

It could be that it's just what the name implies, "South Eurasian". It doesn't necessarily represent an actual population. It just captures similarity to a layer of admixture, including a certain portion of ASI. ASE hits ~50% in Southeast Asians and ~10-15% in Northeast Asians. And "ASI" as we understand it would be a combination of ASE and Oceanian/Papuan in all likelihood.

But why is there this layer of East Eurasian-like admixture that has this relationship to ANE which is spread all over Asia and captures a lot of ASI? I guess it depends on where MA1 is relative to East Eurasian on the tree.

Btw, you can play around with these base components a bit, for example:

https://docs.google.com/spreadsheets/d/1XGY-UIBC2GcUWGl1FizNvu9ofle4ZPHvfNmbrVYxGMo/edit#gid=1118651176

Having a Gedrosian like component (~10% ASE) but dialing down the ANE just a little results in a component whose spread is like an Iranian version of Gedrosian, with South Asian populations almost always opting for the higher ANE components (which conveniently also describes the more ancient components). Wish I hadn't deleted that calculator.

So I think this reinforces Admixture's non-real nature and the best way to approach South Asia might be to reduce it to its component parts (split its East/West/Southeast parts).

Yes I agree. Admixture is not good enough to distinguish between shared and derived ancestry. South Asians as a whole have shared ancestry going back 80000ybp with every population except some Africans (going back 200000 years it is shared with those rest too).

South Asians partly derive their ancestry directly from Africans and partly from East Eurasians (which is ultimately African). It is the descendant of the former that we are seeing in CHG who are related to the Neolithics via Basal.

The East Eurasian (not to be conflated with East Asian seen in NE India) influence is two types - one direct and one which I call wrap-around. The wrap-around one influences NW South Asians more since it came via inner Asia.
South Asians have no WHG of the European type, but they do have something akin to WHG which was present in inner Asia and which came in with this wrap-around East Eurasian which makes South Asians close to eastern Scythians, Alatains, etc. Once the Neolithics separated it appears that they did not again influence South Asia much except at the margins since South Asians show a distinctly lower affinity to EEF.

Anabasis
01-04-2016, 08:56 PM
No, this really is not about Yamnaya being the focal point of many forum members and overshadowing everything else. Simply put, since there is no green component in Yamnaya and there is green in non-CHG carrying Neolithic European samples, then green in the Kumtepe study's ADMIXTURE is not the "teal" typical of CHG.

Well i read what i wrote again and again. i didnt find anything wrong. What i said is simple. The green component is not the CHG component itself but some part of it. I said its CHG LIKE (its a word using to express similaritiy of two thing) component. i also strictly expressed that the theories which based on the CHG ancestry of Yamnaya comes from Kartvelian speaking populations are being collapsed. (Ex: http://eurogenes.blogspot.com.tr/2015/10/yamnayas-exotic-ancestry-kartvelian.html) But i realy hard to understand why he insist on explaining CHG over Yamnaya. We should discuss Kotias and Satsurblia but not Yamnaya while CHG is being argued. On the other hand look what David Says in his last article.

"- Weak segment sharing between the Yamnaya individual and Kotias, a Mesolithic Caucasus hunter-gatherer (CHG) from western Georgia, suggests that the Yamnaya population did not receive its CHG admixture from the southwestern Caucasus."

http://eurogenes.blogspot.com.tr/2016/01/spatio-temporal-segment-sharing.html

So please stoping explaining everything over Yamnaya. I am the last guy want to be descendent of Yamnaya here. I have not obsessions to be descendents of PIE while i am the descendents of the most archiac greek language speakers.

Dr_McNinja
01-04-2016, 08:57 PM
The CHG component here includes "Gedrosian" since Gedrosian populations were also used to source its allele frequencies, and also spillover from a ghost basal CHG-S Indian, hitherto undefined population, which I have discussed in other threads. So don't pay too much attention to it. Instead, focus on ASI, SE Asian and EHG. For Andronovo, ASI was 4.64, and Sintashta it was only 1.02. SE Asian was 0 for both. For MA1, ASI was 19.54. I believe that it is more likely that this ASI in Andronovo and Sintashta would have flowed down from MA1, than have been received from India. AS to whether it is actual ASI, the component modal to Onge, or something similar. It is certainly possible that the ASI in MA1, or even perhaps some of the stuff that is showing up as EHG is from a source basal to both ASI and MA1. The only way to tease it out would be to have a proper ancient sample which is substantially derived from that basal source.There's something odd going on with ANE and South Asians though. They have so much of that MA1-like admixture that they're attracted to any component which, if run through Eurogenes ANE K7 or West Eurasia K8, would have the higher ANE proportion. Almost without fail. It makes me wonder if the ASE/Oceanian don't have a relationship to ANE either. I mean, it seems like a remote chance that we'll recover actual historical "ASI" DNA, so it's our best guess.

My weird calculator results are much less weird when there's no South Asian component which eats up CHG-like admixture. Then I wind up having CHG similar to other people, and just some extra East Asian. But, as seen in your K9 ASI calculator, the missing ANE still shows up in the form of lower EHG and higher WHG relative to others in my area. The WHG is there, but the ANE isn't. It was all "used up". I'd bet if CHG wasn't using up all the ANE, more of our East Asian (represented there by SE-Asian) would have "mated" with it and gone into Siberian.

And then there's the case of MA1 being run through K7 himself giving results which are not far off from Central Asian Indo-Europeans or EHG.

Btw, running the K7/K8 calculator without ANE gives the following:


Me:

14.49% ASE
16.68% WHG
12.16% East_Eurasian
1.69% W-African
3.90% E-African
44.51% ENF
6.57% Oceanian

(My normal Oceanian in this calculator is ~4%)

Sapporo:

13.75% ASE
17.84% WHG
10.58% East_Eurasian
0.62% W-African
3.90% E-African
48.73% ENF
4.57% Oceanian

Haryana (HRP0393)

13.38% ASE
24.29% WHG
10.40% East_Eurasian
1.37% W-African
3.56% E-African
41.55% ENF
5.45% Oceanian

East Asian (Han Chinese IIRC)

29.79% ASE
0.43% WHG
69.26% East_Eurasian
0.01% W-African
0.16% E-African
0.35% ENF
0.00% Oceanian

Sweden

2.30% ASE
69.33% WHG
6.41% East_Eurasian
1.21% W-African
0.82% E-African
17.79% ENF
2.13% Oceanian

Caucasus/Karachay

3.50% ASE
27.63% WHG
10.01% East_Eurasian
1.99% W-African
0.01% E-African
55.26% ENF
1.60% Oceanian

Iranian (DMXX)

5.17% ASE
17.45% WHG
6.60% East_Eurasian
0.96% W-African
2.11% E-African
65.76% ENF
1.95% Oceanian

With ANE:

Me:

28.35% ANE
15.39% ASE
5.51% WHG
4.62% East_Eurasian
0.33% W-African
2.91% E-African
38.78% ENF
4.10% Oceanian

Sapporo:

30.17% ANE
14.43% ASE
6.25% WHG
2.65% East_Eurasian
0.01% W-African
2.23% E-African
42.30% ENF
1.98% Oceanian

HRP0393 Haryana:

29.54% ANE
14.00% ASE
12.97% WHG
2.69% East_Eurasian
0.05% W-African
2.71% E-African
35.33% ENF
2.71% Oceanian

East Asian:

0.30% ANE
29.78% ASE
0.31% WHG
69.19% East_Eurasian
0.01% W-African
0.14% E-African
0.27% ENF
0.00% Oceanian

Sweden:

17.58% ANE
2.76% ASE
63.72% WHG
1.95% East_Eurasian
0.30% W-African
0.32% E-African
12.85% ENF
0.52% Oceanian

(This is the old K7 ENF/WHG which were tweaked in West Eurasia K8, so WHG is overemphasized here)

Caucasus/Karachay

21.61% ANE
3.93% ASE
19.48% WHG
4.31% East_Eurasian
0.48% W-African
0.00% E-African
50.19% ENF
0.01% Oceanian

Iranian

17.88% ANE
5.68% ASE
10.63% WHG
1.90% East_Eurasian
0.20% W-African
1.54% E-African
61.84% ENF
0.33% Oceanian

Tag Heuer
01-04-2016, 09:06 PM
Well problem is in your assumption that the Gedrosian type admixture in Yamnaya is CHG but rest is not. But i insist that we can not nammed it as CHG as far as same component also found in South Caucasian population. Thus we should split CHG into two part. One is South caucasian and the other one is Gedrosia. As far as you dont split it into two different component you can not exclude one as "west asian" but not CHG. Thats my my point.

Why it's South Caucasian, it is Caucasian as it is present among Georgians and Chechens

Dr_McNinja
01-04-2016, 09:09 PM
Yes I agree. Admixture is not good enough to distinguish between shared and derived ancestry. South Asians as a whole have shared ancestry going back 80000ybp with every population except some Africans (going back 200000 years it is shared with those rest too).

South Asians partly derive their ancestry directly from Africans and partly from East Eurasians (which is ultimately African). It is the descendant of the former that we are seeing in CHG who are related to the Neolithics via Basal.

The East Eurasian (not to be conflated with East Asian seen in NE India) influence is two types - one direct and one which I call wrap-around. The wrap-around one influences NW South Asians more since it came via inner Asia. Yeah, totally agree.

South Asians have no WHG of the European type, but they do have something akin to WHG which was present in inner Asia and which came in with this wrap-around East Eurasian which makes South Asians close to eastern Scythians, Alatains, etc. Once the Neolithics separated it appears that they did not again influence South Asia much except at the margins since South Asians show a distinctly lower affinity to EEF.I think the WHG in Haryana Jatts and related North Indian pops is actually from Central Asia, and I think some of it, at least, came from Europe. As David mentioned, there was this surge of WHG in these Scythian-type populations (Andronovo, etc) and it probably was boosted from Europe, not just from EHG.

And in my own opinion based on limited experience, there's just too much shared IBS/IBD. I have larger segments in common with random Europeans than some known family members. And it's always a certain type of European ancestry (presumably where the movement to C-Asia came from, North-Central Europe area). It's not Tatars, or Uyghurs, or other Central/North Asian groups with WHG/East Eurasian. It's not Caucasians either, and not really Russians/Finns either.

I also think the lack of affinity to EEF is because of the strong affinity to ANE (and because of that, EHG). I get plenty of EEF or pure WHG in certain calculators when my lower ANE isn't enough for EHG. Kurd's K9 ASI calculator is an example:

https://docs.google.com/spreadsheets/d/1XGY-UIBC2GcUWGl1FizNvu9ofle4ZPHvfNmbrVYxGMo/edit#gid=1548493524

Look at the EHG:WHG ratio of Haryana Jatts and my dad and I relative to everyone else. That's an example of what I mean by ANE-affinity causing weirdness (also happened in puntdna's new K11 calculator where, after scoring CHG on par with South Indians, I was modeled as 79% Tamil Brahmin, 21% Northwest/North-Central Europe, a very weird combo compared to everyone else, but which averages out to Pakistan area anyway).

EDIT: I wonder if Kurd could run an IBS comparison between N/NW Indians (Brahmins, Jatts) and various WHG containing populations from Uyghurs to North Caucasus to EE/Russia to Europeans and everyone inbetween for whom data is available (including Pamiri Tajiks)...

Anabasis
01-04-2016, 09:13 PM
Why it's South Caucasian, it is Caucasian as it is present among Georgians and Chechens

Yes i know. Thats why i also said "ALSO"

Leeroy Jenkins
01-04-2016, 09:14 PM
Well i read what i wrote again and again. i didnt find anything wrong. What i said is simple. The green component is not the CHG component itself but some part of it. I said its CHG LIKE (its a word using to express similaritiy of two thing) component. i also strictly expressed that the theories which based on the CHG ancestry of Yamnaya comes from Kartvelian speaking populations are being collapsed. (Ex: http://eurogenes.blogspot.com.tr/2015/10/yamnayas-exotic-ancestry-kartvelian.html) But i realy hard to understand why he insist on explaining CHG over Yamnaya. We should discuss Kotias and Satsurblia but not Yamnaya while CHG is being argued. On the other hand look what David Says in his last article.

"- Weak segment sharing between the Yamnaya individual and Kotias, a Mesolithic Caucasus hunter-gatherer (CHG) from western Georgia, suggests that the Yamnaya population did not receive its CHG admixture from the southwestern Caucasus."

http://eurogenes.blogspot.com.tr/2016/01/spatio-temporal-segment-sharing.html

So please stoping explaining everything over Yamnaya. I am the last guy want to be descendent of Yamnaya here. I have not obsessions to be descendents of PIE while i am the descendents of the most archiac greek language speakers.

Sorry that the Yamnaya people make you feel uncomfortable inside and you can't bear the thought of being descended from them, but you are the one that used the Yamnaya peoples to justify your false claim. Not much I can do to help your mental hang up there. However, when you make false claims about the Yamnaya peoples such as:


One more point about the paper that Yamnaya seems not have CHG like ancestry which is described as Dark Green component in paper. Thus the theories related with southern caucasian effect on yamnaya might be collapsed.

Don't be surprised when people try and correct your obvious misunderstanding about the dark green component and Yamnaya peoples; you know, the group of people you used yourself to make your inaccurate point.

Anyways, CHG ancestry may be mixed into the dark blue, dark green and light blue components. It is hard to tell at this point since the paper's ADMIXTURE is so different from what we usually see. Hopefully, when people such as Kurd, David and others get their hands on the Kumtepe sample, we will get better results. However, it is not correct to use only the dark green component as the CHG proxy since it is frequent in non-CHG carrying Neolithic Europeans and the authors themselves claim it is not the equivalent of CHG ancestry; citing the lack of the dark green component in Yamnaya as evidence of its non-CHG affinity.

parasar
01-04-2016, 09:29 PM
... (also happened in puntdna's new K11 calculator where, after scoring CHG on par with South Indians, I was modeled as 79% Tamil Brahmin, 21% Northwest/North-Central Europe, a very weird combo compared to everyone else, but which averages out to Pakistan area anyway).
...

On that I got a surprisingly consistent result!

# Population Percent
1 South_Asian 42.83
2 Caucasus_HG 31.20
3 European_HG 10.30

Using 1 population approximation:
1 Gujarati_A @ 5.285972

Using 2 populations approximation:
1 50% Gujarati_A +50% Gujarati_A @ 5.285972 !!

So Gujarati_A must be close to South_Asian + HG

Anabasis
01-04-2016, 09:51 PM
Sorry that the Yamnaya people make you feel uncomfortable inside and you can't bear the thought of being descended from them, but you are the one that used the Yamnaya peoples to justify your false claim. Not much I can do to help your mental hang up there. However, when you make false claims about the Yamnaya peoples such as:



Don't be surprised when people try and correct your obvious misunderstanding about the dark green component and Yamnaya peoples; you know, the group of people you used yourself to make your inaccurate point.

Anyways, CHG ancestry may be mixed into the dark blue, dark green and light blue components. It is hard to tell at this point since the paper's ADMIXTURE is so different from what we usually see. Hopefully, when people such as Kurd, David and others get their hands on the Kumtepe sample, we will get better results. However, it is not correct to use only the dark green component as the CHG proxy since it is frequent in non-CHG carrying Neolithic Europeans and the authors themselves claim it is not the equivalent of CHG ancestry; citing the lack of the dark green component in Yamnaya as evidence of its non-CHG affinity.

Its not related with being descendent of Yamnaya. My reaction is to some people who thinks that "being yamnaya" as an priviliage just because they think that Yamnaya is root of so called "Aryan Race". Those secret racist idea cause some irrelevant statements and makes people blind an seeing huge picture behind the genetics. As i stated my preious post if we argue CHG we should discuss Kotias but not Yamnaya. Yamnaya never argued as source population of CHG. But those people always discuss Yamnaya almost in every genetic discussin. What isinteresting is that most of those "Yamnaya" fans are not Europeans but mostly from south asian or south central asian. I know the reason of that. But i dont want to tell more.

parasar
01-04-2016, 10:03 PM
Well i read what i wrote again and again. i didnt find anything wrong. What i said is simple. The green component is not the CHG component itself but some part of it. I said its CHG LIKE (its a word using to express similaritiy of two thing) component. i also strictly expressed that the theories which based on the CHG ancestry of Yamnaya comes from Kartvelian speaking populations are being collapsed. (Ex: http://eurogenes.blogspot.com.tr/2015/10/yamnayas-exotic-ancestry-kartvelian.html) But i realy hard to understand why he insist on explaining CHG over Yamnaya. We should discuss Kotias and Satsurblia but not Yamnaya while CHG is being argued. On the other hand look what David Says in his last article.

"- Weak segment sharing between the Yamnaya individual and Kotias, a Mesolithic Caucasus hunter-gatherer (CHG) from western Georgia, suggests that the Yamnaya population did not receive its CHG admixture from the southwestern Caucasus."

http://eurogenes.blogspot.com.tr/2016/01/spatio-temporal-segment-sharing.html

So please stoping explaining everything over Yamnaya. I am the last guy want to be descendent of Yamnaya here. I have not obsessions to be descendents of PIE while i am the descendents of the most archiac greek language speakers.

A very interesting finding by Davidski. I had always suspected Kotias-CHG was not the ANE rich population that merged into EHG.

Though wouldn't you think that archaic Greek speakers were quite Yamna-like?

Leeroy Jenkins
01-04-2016, 10:11 PM
Its not related with being descendent of Yamnaya. My reaction is to some people who thinks that "being yamnaya" as an priviliage just because they think that Yamnaya is root of so called "Aryan Race". Those secret racist idea cause some irrelevant statements and makes people blind an seeing huge picture behind the genetics. As i stated my preious post if we argue CHG we should discuss Kotias but not Yamnaya. Yamnaya never argued as source population of CHG. But those people always discuss Yamnaya almost in every genetic discussin. What isinteresting is that most of those "Yamnaya" fans are not Europeans but mostly from south asian or south central asian. I know the reason of that. But i dont want to tell more.

Fair enough. You have a point that some people relate Yamnaya ancestry with the mythical "Aryan Race" and that Yamnaya discussions and debates tend to find their way into many threads that are not started with the intent of discussing Yamnaya-related topics. I must say, however, that I do not find many people that openly believe or discuss the former on this forum. The latter is certainly a real occurrence, though. I personally only used the Yamnaya peoples as an example because that is the group you and ArmandoR1b were using when I entered the discussion, and there was no racial or ethnical motivation in my posts.

The only point I was trying to make was that the dark green component in this paper's ADMIXTURE output is not the typical "teal" component associated with CHG ancestry and that CHG ancestry may be hidden within multiple components from the output given what we already now about the ancient and modern distributions of CHG ancestry from previous studies.

Gravetto-Danubian
01-04-2016, 10:30 PM
A very interesting finding by Davidski. I had always suspected Kotias-CHG was not the ANE rich population that merged into EHG.

Why would they be ? Kotias are likely an extinct hunter-gatherer. Rather, a different population admixed into Yamnaya which contained CHG, or CHG- like, components.


Though wouldn't you think that archaic Greek speakers were quite Yamna-like?

I seriously doubt it.

Kristiina
01-04-2016, 10:33 PM
I am completely Finnish, and according to Eurogenes K6 my Yamnaya percentage is 0.518909. On the other hand, according to Eurogenes K8, my WHG percentage is 0.498081. I do not know if over half of my genetic ancestry really comes from people who made up the South Russian Yamnaya Culture, but it is easy to believe that my WHG is nearly 50%.

Many Turkic speakers would probably get very high Yamnaya percentages, as well as some Caucasians who speak divergent autochtonous languages.

Therefore, I am sure that the relationship between genetic ancestry and languages/language families is much more complex than often assumed among people who only focus on IE languages.

Gravetto-Danubian
01-04-2016, 10:38 PM
I am completely Finnish, and according to Eurogenes K6 my Yamnaya percentage is 0.518909. On the other hand, according to Eurogenes K8, my WHG percentage is 0.498081. I do not know if over half of my genetic ancestry really comes from people who made up the South Russian Yamnaya Culture, but it is easy to believe that my WHG is nearly 50%.

Many Turkic speakers would probably get very high Yamnaya percentages, as well as some Caucasians who speak divergent autochtonous languages.

Therefore, I am sure that the relationship between genetic ancestry and languages/language families is much more complex than often assumed among people who only focus on IE languages.

Of course it is. But people often look for nice, simple answers to put on their mantle piece.
It is often overlooked that modern Europeans are more like Unetice, whilst Yamnaya -like ancestry is best preserved in circum-Uralic and Turkic speakers.

parasar
01-04-2016, 10:42 PM
Why would they be ? Kotias are likely an extinct hunter-gatherer. Rather, a different population admixed into Yamnaya which contained CHG, or CHG- like, components.

It is not just that the different population contained CHG, or CHG-like, but CHG has very little ANE while the admixing population into the EHG has about just as much ANE as the EHG (about 50%), so much so that after admixture the ANE levels barely changed.




I seriously doubt it.

If CHG=PIE, then I would seriously doubt too. But if Yamna=PIE then I would assume any archaic daughter population of PIE was quite Yamna-like.

Gravetto-Danubian
01-04-2016, 10:46 PM
It is not just that the different population contained CHG, or CHG-like, but CHG has very little ANE while the admixing population into the EHG has about just as much ANE as the EHG (about 50%), so much so that after admixture the ANE levels barely changed.

Yes. So isn't it obvious where we should look ?
:)



If CHG=PIE, then I would seriously doubt too. But if Yamna=PIE then I would assume any archaic daughter population of PIE was quite Yamna-like.

I very much shy against the simplistic doctrines of "X", "Y" or "Z" autosome-components or Y - haplogroups = "PIE", because the expansion of IE must have been a far more complex, protracted and regionally unique affair.

With regard to the 'origins' of BA Greeks, it was multi-faceted, consisting of :

- Baden -derived central European Copper Age components
- Cernadova- Usatovo, viz Cucuteni-Tripolje- components
- CHG via Anatolia
- residual Balkanic EEFs.

(hypothetical construct in lieu of aDNA)

parasar
01-04-2016, 11:01 PM
Yes. So isn't it obvious where we should look ?


It is - among the southern Yamna.




I very much shy against the simplistic doctrines of "X", "Y" or "Z" autosome-components or Y - haplogroups = "PIE", because the expansion of IE must have been a far more complex, protracted and regionally unique affair.

I don't for the most part. Ancient populations were very parochial, and IE ones especially very clannish. We had populations living in India next to each other for thousands of years with very little genetic exchange. The Y-line delineations are especially stark. Sure there are exceptions.



With regard to the 'origins' of BA Greeks, it was multi-faceted, consisting of :

- Baden -derived central European Copper Age components
- Cernadova- Usatovo, viz Cucuteni-Tripolje- components
- CHG via Anatolia
- residual Balkanic EEFs.

Possible. We will see when early IE Greek and Anatolian ancient dna is tested.

Gravetto-Danubian
01-04-2016, 11:06 PM
It is - among the southern Yamna.


Can you please 'stage' this in terms of Meso-, Neo-, Eneolith periods ?

Ryukendo
01-05-2016, 01:26 AM
By ASE I meant the ASE component in Eurogenes K7/K8.

A bunch of these ancient genomes have it, in a pretty ordered gradient. CHG had 5%. The Gedrosian component had 9-10%. Bronze Age Central Asian groups like Afanasievo/Andronovo/Sintashta had it at 8-11%. EHG had it at around 5%. MA1 had 11%.

Modern Caucasus had it at ~4%. Modern Europeans were around 1%.

Actually, in the K7 calculator some of the Afanasievo genomes on Gedmatch (though I think they had low coverage) look like MA1's result with a slight shift towards ENF by 15%.

But why is there this layer of East Eurasian-like admixture that has this relationship to ANE which is spread all over Asia and captures a lot of ASI? I guess it depends on where MA1 is relative to East Eurasian on the tree.

Btw, you can play around with these base components a bit, for example:

https://docs.google.com/spreadsheets/d/1XGY-UIBC2GcUWGl1FizNvu9ofle4ZPHvfNmbrVYxGMo/edit#gid=1118651176

Having a Gedrosian like component (~10% ASE) but dialing down the ANE just a little results in a component whose spread is like an Iranian version of Gedrosian, with South Asian populations almost always opting for the higher ANE components (which conveniently also describes the more ancient components). Wish I hadn't deleted that calculator.

So I think this reinforces Admixture's non-real nature and the best way to approach South Asia might be to reduce it to its component parts (split its East/West/Southeast parts).

I think this pattern is indicative of some kind of allele frequency sharing between MA-1 and S Indians, which is not replicated with EHG and S Indians, as almost all studies, incl. the Haak and Reich original papers where MA-1 was presented, show MA-1 scoring high levels of the 'S Indian' component. Studies where both EHG and MA-1 are run in ADMIXTURE show EHG scoring in 'West Asian/SC Asian' components only, while MA-1 scores in both 'West Asian/SC Asian' and 'S Indian' components.

Furthermore, SC and S Asians are shifted even further towards MA-1 than Kotias, Satsurblia and LNBA Europeans are in a MA-1 vs WHG plot. SC and S Asians are also shifted towards MA-1 in a MA-1 vs Afanasievo plot, deviating away from the trend line. Both of these indicate that the just LNBA+Kotias alone is insufficient to bias SC and S Asians towards MA-1 to the degree that they are.

As for the distribution of the 'S Indian' or affiliated components, I think your own explanation

It could be that it's just what the name implies, "South Eurasian". It doesn't necessarily represent an actual population. It just captures similarity to a layer of admixture, including a certain portion of ASI. .

may suffice (?), but then there is countervailing evidence from Balaji, who posted negative f3 statistics showing an East Asian signal in Middle Easterners, which is still v difficult to explain.

Ryukendo
01-05-2016, 01:44 AM
It is not just that the different population contained CHG, or CHG-like, but CHG has very little ANE while the admixing population into the EHG has about just as much ANE as the EHG (about 50%), so much so that after admixture the ANE levels barely changed.



If CHG=PIE, then I would seriously doubt too. But if Yamna=PIE then I would assume any archaic daughter population of PIE was quite Yamna-like.

Parasar,

Hmm, since I think the evidence is quite solid that S and SC Asia contain some very old ANE ancestry, admixture into Yamnaya from SC and S Asia CHG-related populations would indeed give a boost in the ANE-shift of Yamnaya that admixture from Kotias alone could not have done. However, this depends on that result (that Yamnaya would be insufficiently ANE-shifted if it was a simple two-way mix of CHG and EHG) holding true, and that result was derived from an outdated ADMIXTURE run, where we were still trying to 'suspend' all the ancient samples in between the ANE, WHG, and ENF corners, assuming that everything was a three-way mixture, and we did not take into account the non-admixture/non-clinal aspects of the relationship between the WHG samples, MA-1 and Karelian HG, for example; i.e, there is a certain respect where the split between the WHG genomes, MA-1, and EHG is tree-like, with EHG itself showing no signs of admixture, and we did not take this into account. Not to mention the fact that Kotias itself is probably quite severely mismodelled if we attempt to 'suspend' it between MA-1 and ENF.

Just wondering, because now that we can do direct comparisons with Kotias, Karelian and Yamnaya, the evidence in fact indicates that Yamnaya is in fact *too* ANE-shifted if it is a two-way mixture of Kotias and EHG. Both ADMIXTURE and formal qpAdm modelling, that focus now on each one of the ancient genomes in their own right (instead of the forcible fitting into the triangle we used previously), indicate that Yamnaya is best modelled with an extra dollop of WHG on top of whatever EHG and Kotias ancestry it received. Of course, one can, once again, say that this is because the EHG sample or the Kotias sample is insufficiently WHG-shifted, the samples we have are not the best representatives of the proximal ancestors of the Yamnaya, which are actually deeper into Europe and thus more WHG-shifted, but then this would shift the genomes in the 'wrong' direction, if one is looking for input into Yamnaya from further south or east.

Arbogan
01-05-2016, 02:35 AM
I think it's rather pointless to speculate and discuss regarding the papers results at this point, it's rather superflous. It would be better to simply run the hard data via tools and test some analytical questions see what that yields and how the that confer with hypothesises, explanation models, speculations.

Kurd
01-05-2016, 03:24 AM
However, this depends on that result (that Yamnaya would be insufficiently ANE-shifted if it was a simple two-way mix of CHG and EHG) holding true



POP1
POP2
TARGET
OUTGROUP
DSTAT
Z


Karelia_HG
Yamnaya
MA1
Gorilla
0.0436
4.699


KOTIAS
Yamnaya
MA1
Gorilla
-0.0657
-8.383


KOTIAS
Karelia_HG
MA1
Gorilla
-0.105
-9.852





POP1
POP2
TARGET
OUTGROUP
DSTAT
Z


Kalash
KOTIAS
MA1
Gorilla
0.0317
4.432


Pathan
KOTIAS
MA1
Gorilla
0.0277
4.034


Kalash
Yamnaya
MA1
Gorilla
-0.0342
-6.911


Pathan
Yamnaya
MA1
Gorilla
-0.0388
-8.634





POP1
POP2
TARGET
OUTGROUP
D
Z


Loschbour
Yamnaya
MA1
Gorilla
-0.0317
-3.91


Bichon
Yamnaya
MA1
Gorilla
-0.0298
-3.364


LaBrana1
Yamnaya
MA1
Gorilla
-0.0246
-2.61


Loschbour
KOTIAS
MA1
Gorilla
0.0307
3.124


Bichon
KOTIAS
MA1
Gorilla
0.0343
2.993


LaBrana1
KOTIAS
MA1
Gorilla
0.0402
3.639


GujaratiA
Yamnaya
MA1
Gorilla
-0.0414
-8.218


GujaratiB
Yamnaya
MA1
Gorilla
-0.0443
-8.557


GujaratiC
Yamnaya
MA1
Gorilla
-0.0471
-9.122


GujaratiD
Yamnaya
MA1
Gorilla
-0.0553
-10.269


GujaratiA
KOTIAS
MA1
Gorilla
0.0268
3.661


GujaratiB
KOTIAS
MA1
Gorilla
0.022
2.99


GujaratiC
KOTIAS
MA1
Gorilla
0.0192
2.575


GujaratiD
KOTIAS
MA1
Gorilla
0.0092
1.258


Puliyar
KOTIAS
MA1
Gorilla
-0.0023
-0.289


Puliyar
Yamnaya
MA1
Gorilla
-0.0678
-11.119



EDIT: KOTIAS should not be underestimated. He is almost as MA1 shifted as S Indians

Ryukendo
01-05-2016, 03:57 AM
POP1
POP2
TARGET
OUTGROUP
DSTAT
Z


Karelia_HG
Yamnaya
MA1
Gorilla
0.0436
4.699


KOTIAS
Yamnaya
MA1
Gorilla
-0.0657
-8.383


KOTIAS
Karelia_HG
MA1
Gorilla
-0.105
-9.852





POP1
POP2
TARGET
OUTGROUP
DSTAT
Z


Kalash
KOTIAS
MA1
Gorilla
0.0317
4.432


Pathan
KOTIAS
MA1
Gorilla
0.0277
4.034


Kalash
Yamnaya
MA1
Gorilla
-0.0342
-6.911


Pathan
Yamnaya
MA1
Gorilla
-0.0388
-8.634



Hmmm.... Kurd, what are these results supposed to show? Could you explain?

To me, the stats seem a bit counfounded by Crown vs Basal Eurasian; MA-1 is more similar to Karelia than to Yamnaya, due to more crown Eurasian in Karelian; MA-1 is more similar to Yamnaya than to Kotias, due to more crown Eurasian in Yamnaya; MA-1 is more similar to Yamnaya than to Kalash or Pathan, as Yamnaya is more crown Eurasian and less East Asian, etc.

To answer the question directly, it would be more profitable to do the following I think:
Kotias Yamnaya Loschbour MA1
Primate_Gorilla Karelia_HG Loschbour MA1

Having Kotias and Yamnaya on the same side, and Primate Gorilla and Karelia on the same side, should control for the effects of Crown vs Basal.

Though I am usually against comparing the magnitudes of the D stats, we may try this here just to get some idea. Though we have some nice figures already, as the shared drift of MA-1 and Karelia_HG over Kotias, approximately equals the shared drift of MA-1 and Yamnaya over Kotias plus the shared drift of MA-1 and Karelian over Yamnaya, which tells us that the two halves of Yamnaya must be rather similar to Kotias and Karelian at least in terms of deep ancestry.

Kurd
01-05-2016, 04:16 AM
Hmmm.... Kurd, what are these results supposed to show? Could you explain?

To me, the stats seem a bit counfounded by Crown vs Basal Eurasian; MA-1 is more similar to Karelia than to Yamnaya, due to more crown Eurasian in Karelian; MA-1 is more similar to Yamnaya than to Kotias, due to more crown Eurasian in Yamnaya; MA-1 is more similar to Yamnaya than to Kalash or Pathan, as Yamnaya is more crown Eurasian and less East Asian, etc.

To answer the question directly, it would be more profitable to do the following I think:
Kotias Yamnaya Loschbour MA1
Primate_Gorilla Karelia_HG Loschbour MA1

Having Kotias and Yamnaya on the same side, and Primate Gorilla and Karelia on the same side, should control for the effects of Crown vs Basal.

Though I am usually against comparing the magnitudes of the D stats, we may try this here just to get some idea. Though we have some nice figures already, as the shared drift of MA-1 and Karelia_HG over Kotias, approximately equals the shared drift of MA-1 and Yamnaya over Kotias plus the shared drift of MA-1 and Karelian over Yamnaya, which tells us that the two halves of Yamnaya must be rather similar to Kotias and Karelian at least in terms of deep ancestry.

KOTIAS Yamnaya Loschbour MA1 0.0129 1.684
Loschbour MA1 Karelia_HG Gorilla -0.0058 -0.522

What you say does make sense. Some WHG into the CHG-EHG mix may make a better fit for Yamnaya.

Chad Rohlfsen
01-05-2016, 04:17 AM
POP1
POP2
TARGET
OUTGROUP
DSTAT
Z


Karelia_HG
Yamnaya
MA1
Gorilla
0.0436
4.699


KOTIAS
Yamnaya
MA1
Gorilla
-0.0657
-8.383


KOTIAS
Karelia_HG
MA1
Gorilla
-0.105
-9.852





POP1
POP2
TARGET
OUTGROUP
DSTAT
Z


Kalash
KOTIAS
MA1
Gorilla
0.0317
4.432


Pathan
KOTIAS
MA1
Gorilla
0.0277
4.034


Kalash
Yamnaya
MA1
Gorilla
-0.0342
-6.911


Pathan
Yamnaya
MA1
Gorilla
-0.0388
-8.634





POP1
POP2
TARGET
OUTGROUP
D
Z


Loschbour
Yamnaya
MA1
Gorilla
-0.0317
-3.91


Bichon
Yamnaya
MA1
Gorilla
-0.0298
-3.364


LaBrana1
Yamnaya
MA1
Gorilla
-0.0246
-2.61


Loschbour
KOTIAS
MA1
Gorilla
0.0307
3.124


Bichon
KOTIAS
MA1
Gorilla
0.0343
2.993


LaBrana1
KOTIAS
MA1
Gorilla
0.0402
3.639


GujaratiA
Yamnaya
MA1
Gorilla
-0.0414
-8.218


GujaratiB
Yamnaya
MA1
Gorilla
-0.0443
-8.557


GujaratiC
Yamnaya
MA1
Gorilla
-0.0471
-9.122


GujaratiD
Yamnaya
MA1
Gorilla
-0.0553
-10.269


GujaratiA
KOTIAS
MA1
Gorilla
0.0268
3.661


GujaratiB
KOTIAS
MA1
Gorilla
0.022
2.99


GujaratiC
KOTIAS
MA1
Gorilla
0.0192
2.575


GujaratiD
KOTIAS
MA1
Gorilla
0.0092
1.258


Puliyar
KOTIAS
MA1
Gorilla
-0.0023
-0.289


Puliyar
Yamnaya
MA1
Gorilla
-0.0678
-11.119



EDIT: KOTIAS should not be underestimated. He is almost as MA1 shifted as S Indians

Thanks for the stats. While Kotias may be close to South Asians, South Asians are much further from MA1 than Yamnaya, and will approach significance with Loschbour. The stat that is missing in here, which is very important, is the one which shows that Kotias is not significantly closer to MA1 than Anatolians. Now, how we reconcile that is the question, and maybe you have time to do some digging. Personally, since the beginning I have had a feeling that there is some time of Basal West Eurasian that is more prominent in Kotias than Anatolians, and this will shift them slightly closer to MA1, but artificially. Some type of Basal West Eurasian could also be shifting South Asians towards MA1, if MA1 shares more drift with this group, due to age, than WHG or EHG. This will probably take a lot of digging, and may not be clear without some ancient SC Asians and ASI sample. As far as SC Asians and Central Asians, they do probably contain excess MA1 or EHG stuff. Admixture seems to show this too, as excess EHG, over the rate of WHG to EHG in Sintashta, pops up in SC Asia. qpAdm modeling also works when you use Sintashta, plus MA1 as a pop. It gets harder to model there though, as you know, you can't have two pleft pops that mix to create a third pleft pop. Many good models can be made, it's more a matter of figuring out which makes more sense and agrees with other formal stats.

Chad Rohlfsen
01-05-2016, 04:19 AM
That qpAdm model with some WHG into Yamnaya, does make a much better fit than just straight CHG and EHG.

Ryukendo
01-05-2016, 04:36 AM
Chad, are you proposing a model ((Loschbour, EHG), Basal West Asian---)---------->Kotias and S Asians , with MA-1=Basal West Asian+EHG?

That would indeed explain a lot of the stats.

Chad Rohlfsen
01-05-2016, 04:48 AM
Something to that effect, but MA1 may just be a step along the way to EHG and the group that mixed into Native Americans. MA1 may just by age, share more drift with Basal West Eurasian than WHG and EHG, so it can be modeled as EHG+CHG+minor ENA. I'll run several stats either late tonight or tomorrow. It may have to be tomorrow. I am in the middle of finishing a final paper and I have to be up for work in six hours.

vettor
01-05-2016, 05:00 AM
Sorry that the Yamnaya people make you feel uncomfortable inside and you can't bear the thought of being descended from them, but you are the one that used the Yamnaya peoples to justify your false claim. Not much I can do to help your mental hang up there. However, when you make false claims about the Yamnaya peoples such as:



Don't be surprised when people try and correct your obvious misunderstanding about the dark green component and Yamnaya peoples; you know, the group of people you used yourself to make your inaccurate point.

Anyways, CHG ancestry may be mixed into the dark blue, dark green and light blue components. It is hard to tell at this point since the paper's ADMIXTURE is so different from what we usually see. Hopefully, when people such as Kurd, David and others get their hands on the Kumtepe sample, we will get better results. However, it is not correct to use only the dark green component as the CHG proxy since it is frequent in non-CHG carrying Neolithic Europeans and the authors themselves claim it is not the equivalent of CHG ancestry; citing the lack of the dark green component in Yamnaya as evidence of its non-CHG affinity.

are Gujarati people declared Gedrosian or pure south-asian?

your post #5 shows they have the bulk of sky blue..............is sky blue Only Gedrosian or not

Kurd
01-05-2016, 05:16 AM
I believe Kotias and Indians draw from a W Asian source population basal to both to the exclusion of LBK and Anatolians. Although D stats show Kotias is about as much MA1 shifted as LBK and Anatolians, IBS shows Kotias more MA1 shifted than either Anatolians or LBK. Without a proper reference from this basal source W Asian population, things will not be easily sorted out.

parasar
01-05-2016, 04:12 PM
I believe Kotias and Indians draw from a W Asian source population basal to both to the exclusion of LBK and Anatolians. Although D stats show Kotias is about as much MA1 shifted as LBK and Anatolians, IBS shows Kotias more MA1 shifted than either Anatolians or LBK. Without a proper reference from this basal source W Asian population, things will not be easily sorted out.

Even Dstats show a trend though not statistically significant.
Affinity to MA1 increases geographically west to east and in the Caucasus chronologically as we go away from post-LGM towards the present.
Affinity to MA1: W. Anatolians/SE Europeans < Early Satsurblia CHG < Kotias CHG

So it appears than an MA1 type population was already affecting the eastern areas in the post-LGM and the trend continued thereafter.

It would be interesting to test the hill tribes of Lanka - the Veddas - for their ANE affinity and to see how much ancestry they share with and/or derive from MA1.

As far as South Asia LGM is concerned, in the LGM even NW India became a very dry desert and abandoned and the population was mainly in the south of the Vindhyas. An MA1 type population could have brought ANE to South Asia in the LGM period and it thus managed to penetrate even the southern areas. In the post-LGM the north-west was again reoccupied.

http://www.esd.ornl.gov/projects/qen/nercEURASIA.html

The climate across Indian and south Asia generally seems to have been much more arid than present. Geomorphological indicators from the landmass of India suggest dune mobility in the north-west (Bryason & Swain 1981), and greatly reduced river flow in north-central India during the span of time that covered the full glacial (Williams & Clarke 1984). Offshore indicators of salinity (due to runoff from the land) suggest that LGM aridity was substantially greater than at present. Indicators of upwelling intensity in the Indian Ocean suggest that the summer monsoon was much weaker than present at the LGM, but reaching its weakest at around 15,800 - 12,500 14C years ago, that is 17,800-13,800 calibrated or 'real' years ago (Zonneveld et al. 1997)...

There appears to have been an initial moistening and warming of climate across Eurasia at about 13,000 14C y.a ... From the Bay of Bengal, there are indications that monsoon runoff at 11,500 14C y.a. was in fact greater than today (Cullen 1980), suggesting that rainfall over southern Asia was especially intense (through runoff may also have been aided by the slowness of vegetation recovery).


http://www.esd.ornl.gov/projects/qen/NEW_MAPS/eurasia1.gif

Kurd
01-05-2016, 06:36 PM
I believe Kotias and Indians draw from a W Asian source population basal to both to the exclusion of LBK and Anatolians. Although D stats show Kotias is about as much MA1 shifted as LBK and Anatolians, IBS shows Kotias more MA1 shifted than either Anatolians or LBK. Without a proper reference from this basal source W Asian population, things will not be easily sorted out.

I should have clarified. In my haste, I made it seem that my comments on the basal population, are somehow connected with Kotias's position vs MA1. They are not, instead, my comments are based on ADMIXTURE Dstat and IBS results involving Kotias, Indian tribals, ENF, LBK, and Mota, that wew discussed in other threads. MA1 may have no role with regards to this basal population

ArmandoR1b
01-05-2016, 11:08 PM
Oke i got you. Earth turning around and rise on Yamnaya, Birds sing thier songs for it. Every roads in italy links to Yamnaya, Chiniese built great wall for Yamnaya and CHG is nothing without Yamnaya.

You shouldn't allow your subjectivity to cloud your judgement.

ArmandoR1b
01-05-2016, 11:42 PM
Well i read what i wrote again and again. i didnt find anything wrong. What i said is simple. The green component is not the CHG component itself but some part of it. I said its CHG LIKE (its a word using to express similaritiy of two thing) component. i also strictly expressed that the theories which based on the CHG ancestry of Yamnaya comes from Kartvelian speaking populations are being collapsed. (Ex: http://eurogenes.blogspot.com.tr/2015/10/yamnayas-exotic-ancestry-kartvelian.html) But i realy hard to understand why he insist on explaining CHG over Yamnaya. We should discuss Kotias and Satsurblia but not Yamnaya while CHG is being argued. On the other hand look what David Says in his last article.

"- Weak segment sharing between the Yamnaya individual and Kotias, a Mesolithic Caucasus hunter-gatherer (CHG) from western Georgia, suggests that the Yamnaya population did not receive its CHG admixture from the southwestern Caucasus."

http://eurogenes.blogspot.com.tr/2016/01/spatio-temporal-segment-sharing.html

So please stoping explaining everything over Yamnaya. I am the last guy want to be descendent of Yamnaya here. I have not obsessions to be descendents of PIE while i am the descendents of the most archiac greek language speakers.

CHG was defined by the Kotias and Satsurblia and their DNA was not found in the Early Farmers (other than a minute amount in a few individuals), Middle Neolithic, or Copper Age according to the Jones et al. paper so the Dark Green West Asian is not CHG like. Yamnaya doesn't even have to be used to see this.

http://www.nature.com/ncomms/2015/151116/ncomms9912/fig_tab/ncomms9912_F1.html

ADW_1981
01-06-2016, 01:53 AM
The closest approximation seems to be West Med (orange) and East Med (dark green) labels you see in the Eurogenes calculators.

Arbogan
01-08-2016, 04:32 PM
Can someone .technically savy put kumtepe in D-statistics/qpAdm before there is a war over ancient dead populations. Some analytical questions would be enough to solve the question.

topfield
01-13-2016, 10:49 PM
Not all of Pakistan. Only Iranic and northern Dardic areas. Neighbouring Punjab is an integral part of South Asia.

Dardic areas in Pakistan can be a part of South Asia too due to Dardic people speaking Indic languages. e.g. the Kalash language.

bored
01-14-2016, 12:45 AM
Dardic areas in Pakistan can be a part of South Asia too due to Dardic people speaking Indic languages. e.g. the Kalash language.

Yes, it's debatable. They are genetically closer to their Iranic neighbours though.