PDA

View Full Version : I was wrong - Ancient DNA disproving theories



R.Rocca
11-28-2015, 01:37 PM
While the amount of ancient DNA we have is quite small, and there are still plenty of geographical voids, we have enough of it from Europe to know directionally how Europe was populated. So, I'm starting this thread as an introspective view of my own theories that have been disproved and hope that others share their own. Here's my list:

1. The Lack of R1b in Remedello Culture
I was an early proponent of a Copper Age entry of R1b-L23 into Europe from the steppe, but disagreed with others on the timing. I thought for sure that Remedello would be L51+ due to material culture likeness to Bell Beaker. It is not just that the three Remedello skeletons were all I2a+ that makes this nearly impossible...it is the fact that they plot with other contemporary European Neolithics and lack autosomal DNA that has shifted all modern Europeans, Italians included, towards the steppe.

2. Haplogroup J was not "the" catalyst for Central and Western European farming
I thought this one was a no-brainer, but the dominance of G2a in early farmers showed up in the very first ancient DNA results and hasn't let up. I thought there was still a chance that it would show up in the Balkans, but now we find that Greek Neolithics were G2a as were the great majority of Anatolian Neolithics. Yes, one J2a sample was found in Anatolia, but again, not the overwhelming catalyst I was expecting.

3. E-V13 not a strong player in Central and Western European Farming
This one looked pretty secure when that one Cardial Culture E-V13 sample showed up, but it has been absent in new results since then.

Anabasis
11-28-2015, 08:27 PM
Well about my aDNA theories i was wrong in one thing, and i was right at one as well.

I was wrong that Anatolian Neolethic didnt come up any "teal" (CHG) ancestry which was supposed to be "ANE" ( i was disagree with that defination and thats the what i was right).

I was right that Teal was not step ancestry or something like ANE as far as its disribution shows us the main adress, The Caucasia-Iran pool which seems isolated from it serround during Ice Age untill the neolethic.

Jean M
11-28-2015, 09:14 PM
Everyone can see what needed changing between Ancestral Journeys (2013) and the 2nd edition/paperback (Autumn 2015), so I'll just focus on the big news since then:


Jones 2015. The 4th strand in European ancestry. I had already removed speculation on a Y-DNA J entry into Europe in the Neolithic, but we are now getting a lot closer to understanding this haplogroup. I hope.
Fu 2015. Evidence of interbreeding between Homo sapiens and Neanderthals actually in Europe (most likely) from Peştera cu Oase. I have fought this one. But there it is. Mind you it seems that the Oase population did not contribute substantially to later humans in Europe.

Silesian
11-28-2015, 10:49 PM
Nothing but an exquisite feast of wrong ideas, left me chowing down on humble pie/s !

1) Russell D. Gray and Quentin D. Atkinson http://www.nature.com/nature/journal/v426/n6965/fig_tab/nature02029_F1.html I think JeanM was warning about this one too! Good educational video, https://www.youtube.com/watch?v=4jHsy4xeuoQ

2) R1b-z2103 would show up in Anatolia. R1b-z2103 would not be found anywhere remotely around Yamnaya and related.

3) Age and extent of migration Ydna-IJ- M429[S2/22] and that it would be associated with so many different regions; like WHG+Antolian Farmer+CHG.

In adjusting my wrong line of thinking; proven wrong.
credit to JeanM - time and effort in website and her insight/patience, in explaining or attempting to explain,
& the major blockbuster papers this year,
http://biorxiv.org/content/early/2015/02/10/013433- Haak et al 2015
http://www.nature.com/nature/journal/v522/n7555/full/nature14507.html Allentoft et al 2015
http://www.nature.com/nature/journal/vaop/ncurrent/full/nature16152.html Mathieson et al 2015
http://www.nature.com/ncomms/2015/151116/ncomms9912/full/ncomms9912.html Jones et al 2015
http://www.annualreviews.org/doi/full/10.1146/annurev-linguist-030514-124812 David W. Anthony1 and Don Ringe2 2015

videos on youtube that helped me understand:
https://www.youtube.com/watch?v=4jHsy4xeuoQ
https://www.youtube.com/watch?v=-Q_tqVQHwFw
https://www.youtube.com/watch?v=Z0HCs6PVnzI
https://www.youtube.com/watch?v=d864bwyCAoA
https://www.youtube.com/watch?v=QapUGZ0ObjA
https://www.youtube.com/watch?v=HliaR2Ep24s

nuadha
11-29-2015, 01:01 AM
i didn't think "teal" went back to the ice age.

for me it still remains to be seen if CWC got its genetic signature by admixture within its horizon

if (western) yamnaya largely lacked r1a

if hungarian yamnaya to beaker is really what transformed western europe

One think that still baffles me is the supposed m269 in pre beaker spain.

nuadha
11-29-2015, 01:05 AM
the fact that teal is real and the samara eneolithics contained an r1a does not bode well for me

Gravetto-Danubian
11-29-2015, 01:12 AM
I had always favoured some kind of role for Cucuteni-Tripolye diasporans in the Copper Age events - especially in genesis of CWC
The lack of many findable burials (and that likely C-T were essentially "Balkan-farmer like") doesn't bode well, but I remain optimistic

DMXX
11-29-2015, 01:16 AM
I was entertaining the possibility that Sintashta would possess even more of the "teal"/CHG found in Yamnaya and Afanasievo. I attributed this to the linguistic and archaeological evidence of interactions between Sintashta and the BMAC prior to the Andronovo archaeological horizon's expansion beyond the Urals.

Instead? Sintashta ended up possessing less "teal"/CHG than Yamnaya and contained additional WHG-related admixture. Surprising, but a cautious evidence-based approach has always been my de facto stance. Davidski must still smile regularly in light of how correct his overall perspective was. :)

Still find it amusing I correctly predicted Yamnaya's genetic constitution back in 2012 (see comments (http://vaedhya.blogspot.co.uk/2012/03/north-european-component-variation.html); "In my opinion, the Proto-Indo-European[s] were a West Asian-North European hybrid population ...").

Kale
11-29-2015, 04:07 AM
One think that still baffles me is the supposed m269 in pre beaker spain.

I think we're only going to find more of that in the future. It really looks to me like there was some EHG mixture in Mesolithic Europe. It was suggested as a possibility and modeled in Haak et al. successfully. And with the new Bichon genome out, we can now test it. I think a simple F3 of Bichon,EHG;WHG would make or break it.

Krefter
11-29-2015, 09:29 AM
I first got interested in ancient genetics 3 years ago. That was just after the G2a and Sardinian-autosomal stuff was found in Neolithic, PWC and La Brana being closest to Lithuanian, R1b/R1a associated with IEs because of finds in BBC/CWC/Andronovo. Most of what we learned this year in 2015, was highly suspected because of ancient DNA results.

Most of what I thought back then turned out to be correct. I wasn't surprised at all by the results we got this year. The basics of what I thought about European origins in late 2012 is the same as what i think now. Although a lot of what I thought back then was based on racism and it messed up some of my theories. I couldn't bare the idea of having West Asian ancestry or dark pigmented ancestors, so I'd downplay the role of EEF or Near Eastern ancestry in Yamnaya.

The biggest mistakes I made were.....
Assuming major I clades around today existed in the Mesolithic.
Assuming Eastern HGs were the same as Western HGs.
Assuming Yamnaya was the same as WHG.
Assuming R1a came from Yamnaya.
Assuming Sintashta was 100% Yamnaya.

The core of what I thought of European genetics was....
Sardinian like people expanded in Neolithic
Lithuanian like people expanded in Bronze age.

So, I thought Europeans were mostly a Sardinian+Lithuanian mixture. This is mostly true. Yamnaya wasn't exactly like Lithuanians, but Lithuanians are their closest relatives. And the actually people who brought ANE to North Europe were very similar to Lithuanians. Sintashta, Unetice, and Corded Ware are very similar to Lithuanians.

After Laz 2013 came out and everyone learned about ANE, I was very against the idea ANE in West Asia came from the Steppe. I was also against the idea Basal Eurasian expanded with farming. I always thought ANE in West Asia had a local or South Asian origin and that Basal Eurasian had been around since the Paleolithic. Both these things turned out correct. Although I thought West Asians might have WHG, which turned out wrong.

I was also was against the idea some had that Mesolithic East Europeans were 100% ANE. I thought they had a lot of WHG, which turned out correct. Although I thought they were like 100% WHG. I also doubted Yamnaya had a lot of West Asian ancestry, which turned out correct. I also thought Yamnaya had tons of WHG, which turned out wrong. The WHG comback turned out to be from EEF mixing with WHGs. But Yamnaya has more WHG than most expected.

I was against the idea that Andronovo was very much unlike North Europeans. I knew it wasn't a coincidence they had North European pigmentation, I knew there was some-type of close relationship. I actually suspecting back in November 2012 that Andronovo was not from Yamnaya. I suspected that Yamnaya migrated into Europe mixed with WHGs, then migrated back into Asia with Fair pigmentation they picked up in Europe. At the time I thought that idea was crazy and forgot about it, but it turned out correct.

Krefter
11-29-2015, 09:37 AM
Still find it amusing I correctly predicted Yamnaya's genetic constitution back in 2012 (see comments (http://vaedhya.blogspot.co.uk/2012/03/north-european-component-variation.html); "In my opinion, the Proto-Indo-European were a West Asian-North European hybrid population ...").

Wow!! Dienkies was also a bit of a prophet by predicting the West Asian-signal in Europe came with Indo European languages.

Lank
11-29-2015, 12:11 PM
I found it hard to believe that Basal Eurasian could have remained isolated in the Near East for tens of thousands of years. So, I expected Basal Eurasian to represent a relatively recent migration from Africa, presumably from the North (to explain the 'more divergent than Crown Eurasians, but less divergent than Africans' pattern), and associated BE with the influx of Y-DNA E1b1b1, perhaps during the Natufian period. This scenario was actually added to the preprint of the original paper after we had been discussing it online.

But it turned out to be wrong. We now have a Paleolithic genome from the Caucasus, with no less Basal than the early farmers.

lgmayka
11-29-2015, 12:13 PM
Dienkies was also a bit of a prophet by predicting the West Asian-signal in Europe came with Indo European languages.
Yes, he noticed the Caucasian signal, but for the longest time he denied that the Proto-Indo-Europeans could be connected with the Pontic Steppe, R1a, or R1b.

Dienekes 7/15/2012 (http://dienekes.blogspot.com/2012/07/hints-of-eastcentral-asian-admixture-in.html)
---
My own working hypothesis would derive the earliest Proto-Indo-Europeans with groups living in Neolithic eastern Anatolia and northern Mesopotamia.
---

Dienekes 7/19/2012 (http://dienekes.blogspot.com/2012/07/huge-study-on-y-chromosome-variation-in.html)
---
This is due to the fact that neither R1a nor R1b were originally part of the Indo-European community, but their geographical position was such that they came under the influence of the Indo-Europeans when the latter began their expansion.
---

R.Rocca
11-29-2015, 02:03 PM
I had always favoured some kind of role for Cucuteni-Tripolye diasporans in the Copper Age events - especially in genesis of CWC
The lack of many findable burials (and that likely C-T were essentially "Balkan-farmer like") doesn't bode well, but I remain optimistic

Good one G-D, I had forgotten about Cucuteni-Tripolye. I thought that there was a chance that western-most CT was some form of farmer plus R1b mix and that eastern CT was some form of R1b-R1a mix. If I had to guess now, I'd say there were Neolithic I2a-revival types. Unfortunately, it looks like we will only be able to infer who they were, since they didn't bury their dead.

DMXX
11-29-2015, 03:53 PM
Wow!! Dienkies was also a bit of a prophet by predicting the West Asian-signal in Europe came with Indo European languages.

The only thing I could be described as being is observant. :P

I maintained for several years that both Davidski and Dienekes were both technically "right" in a sense; Dienekes' conclusion that a West Asian signal correlated with the Indo-Europeans on the one hand in Europe, and Davidski's conclusion that a European signal correlated with IE's on the other in Asia, was essentially set up as a false dichotomy. I highlighted this at the end of 2012 in a thread here at Anthrogenica (will edit link in if I find it... Here it is! (http://www.anthrogenica.com/showthread.php?568-Inferences-on-PIE-genetic-make-up-via-modern-trends)).

As we now know from aDNA, Davidski tracked EHG and Dienekes CHG through the cryptic mist of modern DNA before either ancient component were discovered.

alan
11-29-2015, 04:17 PM
There are so many theories I have dabbled in it is inevitable that many are wrong and some are right. One of the problems in my first year or two in this hobby was believing the geneticists dating conclusions and using them as an anchor point to consider the archaeological possibilities. I was guilty of underestimating just how epically wrong the genetic dating based on modern populations could be. It seems very weird to me in hindsight how the geneticists didnt predict that using R1b as a collective unrefined block to date its appearance in Europe wouldnt grossly inflate the age. Also the 3 times fudge factor thing. However once I had a bit of interaction with Ken N about variance I quickly concluded that it was nonsense that R1b was 30000 years old in Europe if a nearly European-only clade like yI had five times as much variance. Europe was only permanently settled about 40000BC. I figured based on that that European R1b, or ht15 as it was then sort of known, couldnt be any older than 8000 years old because because 40000 divided by five.

I suppose going back a lot of years, the idea that yDNA lines of the beaker people could come to dominate western Europe seemed highly improbable to archaeologists given what they had been taught. One thing that I always thought went against a Neolithic farmer origin for most Euro yDNA is the fact that the high R1b zone seemed to ignore/cut across the divide between the north/central European Neolithic and the Cardial Neolithic. However it seems that my skepticism that R1b of the main European type spread with farmers was the right conclusion for the wrong reason now that it appears both north and south Neolithic waves were so similar. I also spotted fairly early (but by no means the first) that beaker and high European R1b seemed to have a geographical correlation. However, the idea that it could have actually been causal was counterintuitive to what I had been taught. What I had been taught was it probably was a migration (thankfully I wasnt taught but arch anti-migrationists) but a very small scale one with influence way beyond its likely genetic impact. However, that said, I was also aware of the massive top-down expansion that clan societies like Gaelic Ireland demonstrated for yDNA so it never seemed totally impossible.

I suppose as ancient DNA came out that I tended to look to R1b as likely being on the fringes of a largely R1a steppe - perhaps in the Caucasus. R1a fantatics were very bad for browbeating and seeming triumphalist about R1a appearing first in ancient DNA samples across the steppe and it slowly made me less optimistic that R1b would be found in Yamnaya itself. Almost all eastern Yamnaya being 22103 was a big surprise. However we had by then become aware of the amazing way yDNA proportions can utterly change. What it did was put the nail in the coffin of using modern populations to project too far back. IMO modern population yDNA studies mostly tell you about the last 2000 years, little certain about 4-5000 years ago and almost nothing about 8000 years ago+

alan
11-29-2015, 04:22 PM
Wow!! Dienkies was also a bit of a prophet by predicting the West Asian-signal in Europe came with Indo European languages.

As with most things a lot of people got it partly right and partly wrong. Its like in disputes in life, often turns out everyone is partly right

alan
11-29-2015, 04:28 PM
I should probably add that if ever the early CHG hunter gatherer from 12000BC?? is shown to have an ANE component in its ancestry then my theory of ANE being bottled up around Altai and south-central Asia fringes (until a move west is indicated by the spread of pressure flake microbles west after 9500BC) is in trouble. However I dont know what to make of the autosomal conclusions as everyone seems to be contradicting each other - am waiting for the dust to settle on CHG.

George
11-29-2015, 04:43 PM
Good one G-D, I had forgotten about Cucuteni-Tripolye. I thought that there was a chance that western-most CT was some form of farmer plus R1b mix and that eastern CT was some form of R1b-R1a mix. If I had to guess now, I'd say there were Neolithic I2a-revival types. Unfortunately, it looks like we will only be able to infer who they were, since they didn't bury their dead.

Some CT aDna might still be recoverable, e.g. the early stages of the eventual Sofiivka type in the north, and of the Vykhvatynsk groups which fused with steppe peoples to create Usatovo (these would be the most interesting), and of the Westernmost CT migrants who transmogrified into Zhyvotylivka (which was among the westernmost segments of Poststog). Another very Western Trypilian group (Gradeshti) had mixed burial rites, so something might also be recovered there. But the classic Central Trypilia groups of the mega-protocities are unfortunately lost to us. We can only speculate about their genetics on the basis of archaeological sequences (which are clearly insufficient). There are also the females of the eastern GAC groups, which appear to be largely of "local" (=Trypilian) origin (spoils of war...).

Krefter
11-29-2015, 06:03 PM
I should probably add that if ever the early CHG hunter gatherer from 12000BC?? is shown to have an ANE component in its ancestry then my theory of ANE being bottled up around Altai and south-central Asia fringes (until a move west is indicated by the spread of pressure flake microbles west after 9500BC) is in trouble. However I dont know what to make of the autosomal conclusions as everyone seems to be contradicting each other - am waiting for the dust to settle on CHG.

If West Asians have ANE then CHG had ANE. The ANE signal in West Asia looks like it has mostly a CHG source. Although the whole idea of ANE may change. Actual descendants of MA1 like people might only be Native Americans, Siberians, EHG, and South Asians.

alan
11-29-2015, 06:17 PM
Some CT aDna might still be recoverable, e.g. the early stages of the eventual Sofiivka type in the north, and of the Vykhvatynsk groups which fused with steppe peoples to create Usatovo (these would be the most interesting), and of the Westernmost CT migrants who transmogrified into Zhyvotylivka (which was among the westernmost segments of Poststog). Another very Western Trypilian group (Gradeshti) had mixed burial rites, so something might also be recovered there. But the classic Central Trypilia groups of the mega-protocities are unfortunately lost to us. We can only speculate about their genetics on the basis of archaeological sequences (which are clearly insufficient). There are also the females of the eastern GAC groups, which appear to be largely of "local" (=Trypilian) origin (spoils of war...).

However, stray human bones turn up in small nos in settlement sites too even if formal burial are elusive. I dont know if this is the case in this example but it often is true.

Romilius
11-29-2015, 06:18 PM
I also was wrong: I would have put my right hand on a fire when I thought that the Caucasian input in Yamnaya was from R1b men from the South.

alan
11-29-2015, 06:18 PM
I like this thread. Being humble is IMO a very important part of being a decent person.

rms2
11-29-2015, 07:26 PM
I expected more y-dna E and J among European Neolithic remains. There is that E-V13 from the Avellaner site in Spain, and the early Lengyel E-M78 and J2-M172 finds from Hungary, but I thought there would be more of that and thus more of a balance between those and G2a and I2a among Neolithic farmers.

MitchellSince1893
11-29-2015, 07:43 PM
Based on present day distribution maps, I was of the opinion that the birthplace for P312 and it's subclades U152, L21, DF27 was probably north of the Alps in Southern Germany. Recent ancient DNA evidence has me thinking it's going to be further East. The question is how far East. Czech Republic? Hungary? Romania? Moldova? Ukraine? Pontic Steppe?

No ancient DNA evidence yet, but based on present day results (little to no U152 in the Belgae Civitas, in present day Isle of Wight and Hampshire), I'm questioning my belief that the Belgae were a major source for U152 in Britain.

Also, based on present day distributions I'm thinking the Iceni tribe was heavy in U152.

I'll have to wait for ancient evidence to nail all of these down one way or the other.

rms2
11-29-2015, 08:13 PM
I like this thread. Being humble is IMO a very important part of being a decent person.

Unless humility is the thing one is proudest of. ;)

George
11-29-2015, 08:22 PM
However, stray human bones turn up in small nos in settlement sites too even if formal burial are elusive. I dont know if this is the case in this example but it often is true.

There may indeed be some additional stray bones in the "adequate" settlements mentioned (i.e. such that were accompanied by relevant inhumation areas). But the overwhelming number of classical Trypilian settlement sites were ritually burned in hugely powerful conflagrations leaving nothing but "ploshchadkas" (incinerated "squares"), which yield material for archaeologists but not for geneticists. No one has yet provided an adequate explanation of the standard burial rite of classical Trypilia. One is free to keep on guessing. Strangely enough, the formal Encyclopedia of Trypilia (Kyiv, 2004, 2 vols.) does not even discuss the issue, in spite of a lengthy and very interesting chapter on Trypilian spirituality: http://trypillia.com/images/encyclopedia/t1-ch5.pdf

Piquerobi
11-29-2015, 08:27 PM
Years ago, I thought the Neanderthal man would have played a for more important role as the ancestor of Europeans. I was quite surprised no mtDNA or yDNA Neanderthal haplogroup turned out among contemporary humans. It still surprises me no archaic mtDNA or yDNA have showed up.

I also got surprised by how Neolithic Europeans were genetically so similar to each other, plotting in a similar place and their similarity to contemporary Sardinians (I would not have guessed it). I am still surprised at how important they are as ancestors to contemporary Europeans.

I was kind of skeptical IE expansions involved genetic shifts. Only when I realized the massive R1b in Western Europe and R1a and India, did I admit IE expansion was more than just adopting a "new fashion". It involved some level of conquest and genetic autosomal input, besides replacement of paternal lineages in a large scale. I understood this when I learnt R1b-M269 was just a recent arrival in Western Europe. The timeframe matched the arrival of IE languages, and also the East-West movement.

lgmayka
11-29-2015, 08:36 PM
However we had by then become aware of the amazing way yDNA proportions can utterly change. What it did was put the nail in the coffin of using modern populations to project too far back. IMO modern population yDNA studies mostly tell you about the last 2000 years, little certain about 4-5000 years ago and almost nothing about 8000 years ago+
I submit that modern yDNA is highly informative even to many thousands of years back; but as you say, the proportions can change drastically. The largest percentages today are often most likely to be fairly recent arrivals, whereas earlier migrations are often represented only by mysterious remnants. For a long time my favorite example of a remnant was N-Y6503 (http://yfull.com/tree/N-Y6503/), which academic publications still do not recognize even when it appears both in ancient and modern European DNA (and nowhere else); but more recently, I have also taken a liking to Q-YP1669 (http://yfull.com/tree/Q-YP1669/), whose (exclusively?) European presence is also hard to explain.

But I will join in the "confessions" by admitting that I was shocked and dismayed when the first aDNA from Yamnaya turned out to be not R1a but R1b. :) And I also viscerally resisted the notion that Sardinians could have their own 4500-year-old clade of R-M458 (namely R-PF6188) (http://yfull.com/tree/R-M458/).

Padre Organtino
11-29-2015, 08:52 PM
I thought Georgians were very much an isolated Neolithic farmer group which is obviously not true as we can see now. I also though that there was a strong Caucasus-like admixture in the Balkans which is again not the case in reality.
Plus my "common sense" understanding of the Russian ancestry involved Russians having far more Finno-Ugric admixture than they have in reality.
Additionally I thought that Basques and Sardinians were distantly related to modern Anatolia and Caucasus which is again false.

ArmandoR1b
11-29-2015, 09:43 PM
As we now know from aDNA, Davidski tracked EHG and Dienekes CHG through the cryptic mist of modern DNA before either ancient component were discovered.

This is an extremely important point. Not just to give due respect but also because it needs to be taken into consideration if we look back on the older posts trying to determine what the results and analysis by each of them meant.

I have always been a fan of Dienekes, and I still am, but when I first read that he did not think that R1b had anything to do with IE in Europe, at least western Europe, I strongly felt that he would be proven wrong on that count. The evidence continues to point that way and we are just waiting for the final death blow on alternate hypotheses.

parastais
11-29-2015, 09:45 PM
A lot of surprises. Now I am not speculating at all :)
OK, I am still, a bit, now and then...

DMXX
11-29-2015, 09:56 PM
This is an extremely important point. Not just to give due respect but also because it needs to be taken into consideration if we look back on the older posts trying to determine what the results and analysis by each of them meant.


Absolutely. Retrospection is crucial if we are to avoid making the same mistakes and approach the remaining unknowns in a manner which prevents unnecessary speculative detours or preoccupations along the way.

May we continue to learn from our collective past errors.

R.Rocca
11-30-2015, 01:11 AM
But I will join in the "confessions" by admitting that I was shocked and dismayed when the first aDNA from Yamnaya turned out to be not R1a but R1b. :) And I also viscerally resisted the notion that Sardinians could have their own 4500-year-old clade of R-M458 (namely R-PF6188) (http://yfull.com/tree/R-M458/).

I was shocked as well, I thought for sure Yamnaya would be almost entirely R1a and that prior steppe cultures like Kemi Oba would have been R1b types that got pushed westward. It will be interesting to see how exactly it is that R1a moved on to become Corded Ware's biggest player and how R1b-L51 came to be differentiated from its R1b-Z2103 brother clade.

royking
11-30-2015, 01:51 AM
I first got interested in ancient genetics 3 years ago. That was just after the G2a and Sardinian-autosomal stuff was found in Neolithic, PWC and La Brana being closest to Lithuanian, R1b/R1a associated with IEs because of finds in BBC/CWC/Andronovo. Most of what we learned this year in 2015, was highly suspected because of ancient DNA results.

Most of what I thought back then turned out to be correct. I wasn't surprised at all by the results we got this year. The basics of what I thought about European origins in late 2012 is the same as what i think now. Although a lot of what I thought back then was based on racism and it messed up some of my theories. I couldn't bare the idea of having West Asian ancestry or dark pigmented ancestors, so I'd downplay the role of EEF or Near Eastern ancestry in Yamnaya.

The biggest mistakes I made were.....
Assuming major I clades around today existed in the Mesolithic.
Assuming Eastern HGs were the same as Western HGs.
Assuming Yamnaya was the same as WHG.
Assuming R1a came from Yamnaya.
Assuming Sintashta was 100% Yamnaya.

The core of what I thought of European genetics was....
Sardinian like people expanded in Neolithic
Lithuanian like people expanded in Bronze age.

So, I thought Europeans were mostly a Sardinian+Lithuanian mixture. This is mostly true. Yamnaya wasn't exactly like Lithuanians, but Lithuanians are their closest relatives. And the actually people who brought ANE to North Europe were very similar to Lithuanians. Sintashta, Unetice, and Corded Ware are very similar to Lithuanians.

After Laz 2013 came out and everyone learned about ANE, I was very against the idea ANE in West Asia came from the Steppe. I was also against the idea Basal Eurasian expanded with farming. I always thought ANE in West Asia had a local or South Asian origin and that Basal Eurasian had been around since the Paleolithic. Both these things turned out correct. Although I thought West Asians might have WHG, which turned out wrong.

I was also was against the idea some had that Mesolithic East Europeans were 100% ANE. I thought they had a lot of WHG, which turned out correct. Although I thought they were like 100% WHG. I also doubted Yamnaya had a lot of West Asian ancestry, which turned out correct. I also thought Yamnaya had tons of WHG, which turned out wrong. The WHG comback turned out to be from EEF mixing with WHGs. But Yamnaya has more WHG than most expected.

I was against the idea that Andronovo was very much unlike North Europeans. I knew it wasn't a coincidence they had North European pigmentation, I knew there was some-type of close relationship. I actually suspecting back in November 2012 that Andronovo was not from Yamnaya. I suspected that Yamnaya migrated into Europe mixed with WHGs, then migrated back into Asia with Fair pigmentation they picked up in Europe. At the time I thought that idea was crazy and forgot about it, but it turned out correct.

Krefter, I admire your courage and honesty. As a person of color I celebrate admixture and hybridity, so I am keen to see your evolution! Yet, even though I don't see any particular genetic biases, personally, I was completely wrong at least 10-12 years ago when I proposed that J2 and E tracked the initial Neolithic into Europe. Haplogroup G frequency did not correlate with the presence of painted pottery and figurines, mostly because Georgia had so much G compared to parts of Europe, it dwarfed the association. We had always assumed some genetic continuity in the Y chromosome across millennia. No one had predicted that R1b and R1b were post-Neolithic expansions, nor that J2a mostly arrived in Europe in the Late Neolithic to Bronze Ages. The beauty and awe of aDNA studies!

Agamemnon
11-30-2015, 02:53 AM
Don't laugh: I initially thought G2a had something to do with the spread of IE, of course that was before we had genomic data from Neolithic Europe, but still, that was a pretty silly theory alright. On the other hand, I always had a hard time accepting models which tied haplogroup J to the spread of agriculture, I remember reading Roy King's 2009's article entitled "Neolithic Migrations in the Near East and the Aegean" (a pretty fascinating article) and remaining skeptical ever since. J2a's appearance in Bronze Age Hungary wasn't much of a surprise to me because I always thought (and still think) that J2a's current frequencies are the result of major Bronze Age expansions. I also thought E-V13 was "born" in Northeast Africa, now that we know more about M78's phylogeny the odds really aren't in favour of this theory. Finally, I once thought I1 had been in Scandinavia since the Mesolithic, needless to say I didn't expect I1 to make its first appearance in a Hungarian LBK sample, nor did I expect I2a and I2c to show up instead of I1 in Scandinavia back during the Mesolithic.

Gravetto-Danubian
11-30-2015, 03:27 AM
I was completely wrong at least 10-12 years ago when I proposed that J2 and E tracked the initial Neolithic into Europe. Haplogroup G frequency did not correlate with the presence of painted pottery and figurines, mostly because Georgia had so much G compared to parts of Europe, it dwarfed the association. We had always assumed some genetic continuity in the Y chromosome across millennia. No one had predicted that R1b and R1b were post-Neolithic expansions, nor that J2a mostly arrived in Europe in the Late Neolithic to Bronze Ages. The beauty and awe of aDNA studies!

Roy, let me say that although proven incorrect, your hypotheses helped shape what we know today. You and your team are one of the pioneers, and at least had credible crack at putting a hypothesis together. Without the early hypotheses based on modern data, people would not have been led to in turn question whether genetic patterns really do underlie movements, and whether they really can be linked to cultural phenomena.

I know critiques of genetic anthropology said the patterns are just noise; or indecipherable palimpsests; and pots (or Neolithic figurines) don't equal people. Whilst some aspects wer correct; the overly immobilist views have themselves been utterly falsified

Agamemnon
11-30-2015, 03:45 AM
Roy, let me say that although proven incorrect, your hypotheses helped shape what we know today. You and your team are one of the pioneers, and at least had credible crack at putting a hypothesis together. Without the early hypotheses based on modern data, people would not have been led to in turn question whether genetic patterns really do underlie movements, and whether they really can be linked to cultural phenomena.

I second that entirely.

Arbogan
11-30-2015, 04:43 AM
Gedrosia was not caucasus related, but south-central asian/Iran related.

parasar
11-30-2015, 05:00 AM
...
Assuming major I clades around today existed in the Mesolithic.
Assuming Eastern HGs were the same as Western HGs.
Assuming Yamnaya was the same as WHG.
Assuming R1a came from Yamnaya.
Assuming Sintashta was 100% Yamnaya. ...


"Assuming major I clades around today existed in the Mesolithic." They did.
"Assuming Eastern HGs were the same as Western HGs." It's a continuum. You can always pick up discreet units from the continuum that will be different.
"Assuming R1a came from Yamnaya." If you mean R1a-M417, it likely did.
"Assuming Sintashta was 100% Yamnaya." Sintashta is an offshoot of Yamna. Offshoots can be thought of as 100% descendants but still may be a small proportion of the parent population's diversity.

parasar
11-30-2015, 05:46 AM
Don't laugh: I initially thought G2a had something to do with the spread of IE, of course that was before we had genomic data from Neolithic Europe, but still, that was a pretty silly theory alright ...

G2a today has an India to Europe distribution. So would you say that G2a hitched onto the R1 IE bandwagon? I do think that IE was spread by R1 (and R2), but it is not clear whether PIE originated within R.

Michał
11-30-2015, 09:28 AM
I made far too many wrong predictions to list them all here, so let me focus on those results that surprised me the most.

1. Back in 2006, I expected Corded Ware to be associated mostly with R1b (more specifically with U106), so the Eulau results (2008) were a kind of surprise. The presence of R1b in CWC is still an option, but we know by now that R1a was by far the dominant haplogrup there.
2. I was extremely skeptical regarding the possible Neanderthal admixture in modern humans until the ancient DNA data definitely proved it really happened (more than once!).
3. While I was strongly convinced that Yamna will show R1b-Z2103, I also expected R1a-Z93 to be found in Eastern Yamna, which turned out to be totally wrong.

kingjohn
11-30-2015, 09:41 AM
i was wrong about my haplogroup e-m123-m34
i always thought it was limited to 3 places : ethiopia, israel, dead sea jordanian.
but than came this paper http://www.nature.com/nature/journal/v522/n7555/full/nature14507.html
which found rise423 in bronze age armenia i was shocked i always thought it is limited to much southern areas.
adam

MfA
11-30-2015, 09:50 AM
i was wrong about my haplogroup e-m123-m34
i always thought it was limited to 3 places : ethiopia, israel, dead sea jordanian.
but than came this paper http://www.nature.com/nature/journal/v522/n7555/full/nature14507.html
which found rise423 in bronze age armenia i was shocked i always thought it is limited to much southern areas.
adam

Jewish M34 is almost always
M34>M84>Y14899
M34>L791>Y6923

kingjohn
11-30-2015, 10:19 AM
dear Mfa what is your thought on the presence of this e-m34
that north in bronze age armenia ?
were you also surprised?
p.s
those you mention are aschenazi clusters my direct paternal line is mizrachi
regards
adam

lgmayka
11-30-2015, 10:31 AM
"Assuming major I clades around today existed in the Mesolithic." They did.
Perhaps he is referring to some surprisingly young TMRCAs.
- I1, 4700 ybp (http://yfull.com/tree/I1/)
- I-L161, 6900 ybp (http://yfull.com/tree/I-L161.1/)
- I-L621, 6500 ybp (http://yfull.com/tree/I-L621/)

So if we picture the world 7000 years ago, three specific men living at that time sired all the I1, I-L161, and I-L621 men living today (patrilineally speaking).

Or put another way: The I1, I-L161, and I-L621 clades as we know them today did not exist 7000 years ago.

MfA
11-30-2015, 10:33 AM
dear Mfa what is your thought on the presence of this e-m34
that north in bronze age armenia ?
were you also surprised?
p.s
those you mention are aschenazi clusters my direct paternal line is mizrachi
regards
adam

You're right they're Ashkenazi clusters, but there is a Palestinian Christian guy in M34>M84>Y14899 cluster who could be related to Mizrahi Jews?
I don't have a strict opinion honestly about M84+>PF6751- MBA Armenian. The dates are younger than the Mesopotamian Semites, so could be related to them or not.

Gravetto-Danubian
11-30-2015, 10:37 AM
Perhaps he is referring to some surprisingly young TMRCAs.
- I1, 4700 ybp (http://yfull.com/tree/I1/)
- I-L161, 6900 ybp (http://yfull.com/tree/I-L161.1/)
- I-L621, 6500 ybp (http://yfull.com/tree/I-L621/)

So if we picture the world 7000 years ago, three specific men living at that time sired all the I1, I-L161, and I-L621 men living today.

I find this still an intriguing concept . As we all know, these "super -fathers" were but one man amongst many kin in their tribe. So that only one passed on his specific line must have necessitated, both, a large element of chance and inevitability - for some reason Y lines behave that way ?

rms2
11-30-2015, 01:05 PM
. . . No one had predicted that R1b and R1b were post-Neolithic expansions . . .

Actually, a few of us did. I was arguing on Rootsweb, at FTDNA's forum, and on the old dna forums for that and the connection of both R1b and R1a to the early Indo-Europeans as long ago as 2006. Of course, I am neither famous nor a professional scientist, so I was mostly laughed off.

parasar
11-30-2015, 03:26 PM
Perhaps he is referring to some surprisingly young TMRCAs.
- I1, 4700 ybp (http://yfull.com/tree/I1/)
- I-L161, 6900 ybp (http://yfull.com/tree/I-L161.1/)
- I-L621, 6500 ybp (http://yfull.com/tree/I-L621/)

So if we picture the world 7000 years ago, three specific men living at that time sired all the I1, I-L161, and I-L621 men living today (patrilineally speaking).

Or put another way: The I1, I-L161, and I-L621 clades as we know them today did not exist 7000 years ago.

True, a number of papers such as "Y Chromosomes of 40% Chinese Are Descendants of Three Neolithic Super-grandfathers" and "Modeling the contrasting Neolithic male lineage expansions in Europe and Africa" have pointed out relatively recent Y line expansions.

Nevertheless we see I1 shared SNPs in specimens both from Hungarian LBK (http://biorxiv.org/content/biorxiv/early/2014/09/03/008664.full.pdf) and Mesolithic Sweden (Stora Forvar 11) so while pruned, related lines appear to be quite old in Europe.

parasar
11-30-2015, 03:50 PM
I made far too many wrong predictions to list them all here, so let me focus on those results that surprised me the most.

1. Back in 2006, I expected Corded Ware to be associated mostly with R1b (more specifically with U106), so the Eulau results (2008) were a kind of surprise. The presence of R1b in CWC is still an option, but we know by now that R1a was by far the dominant haplogrup there.
2. I was extremely skeptical regarding the possible Neanderthal admixture in modern humans until the ancient DNA data definitely proved it really happened (more than once!).
3. While I was strongly convinced that Yamna will show R1b-Z2103, I also expected R1a-Z93 to be found in Eastern Yamna, which turned out to be totally wrong.

Michał,

Why do you think you were totally wrong on #3? After all 4.9-4.7ka sample I0432 from the Samara Bend is Z94+

I confess my thinking that M417 was born near the North Sea/English Channel (based on modern distribution of CTS4385) has not panned out.

Michał
11-30-2015, 06:19 PM
Michał,
Why do you think you were totally wrong on #3? After all 4.9-4.7ka sample I0432 from the Samara Bend is Z94+

I was actually expecting R1a-Z93 to be a dominant haplogroup in Eastern Yamna and in the descending Poltavka culture, while it turned out that Z93 dominated the steppe much later and was initially a part of the broad Corded Ware Horizon (or its Eastern part that produced the Abashevo culture, ancestral to Sintashta), or at least this is what is quite strongly suggested by the autosomal results of the Sintashta samples. This association with Corded Ware and Abashevo was something I considered a possible but, nevertheless, less likely scenario (http://www.anthrogenica.com/showthread.php?1516-R1a-and-Corded-Ware&p=17642&viewfull=1#post17642).

parasar
11-30-2015, 07:30 PM
But I was actually expecting R1a-Z93 to be a dominant haplogroup in Eastern Yamna and in the descending Poltavka culture, while it turned out that Z93 dominated the steppe much later and was initially a part of the broad Corded Ware Horizon (or its Eastern part that produced the Abashevo culture, ancestral to Sintashta), or at least this is what is quite strongly suggested by the autosomal results of the Sintashta samples. This association with Corded Ware and Abashevo was something I considered a possible but, nevertheless, less likely scenario (http://www.anthrogenica.com/showthread.php?1516-R1a-and-Corded-Ware&p=17642&viewfull=1#post17642).

It's my feeling that folk are reading the autosomal results of Sintashta wrongly.

Mathieson et al. did slightly walk away from Allentoft et al.'s suggestion:
"Previous work documented that such ancestry appeared east of the Urals beginning at least by the time of the Sintashta culture, and suggested that it reflected an eastward migration from the Corded Ware peoples of central Europe5. However, the fact that the Srubnaya also harbored such ancestry indicates that the Anatolian Neolithic or EEF ancestry could have come into the steppe from a more eastern source."

But IMO this is not sufficient; as Sample I0432 also had such ancestry it should cause a complete rethinking of autosomal relationships of Corded Ware vis a vis Sintashta and Andronovo.

Coldmountains
11-30-2015, 08:11 PM
It's my feeling that folk are reading the autosomal results of Sintashta wrongly.

Mathieson et al. did slightly walk away from Allentoft et al.'s suggestion:
"Previous work documented that such ancestry appeared east of the Urals beginning at least by the time of the Sintashta culture, and suggested that it reflected an eastward migration from the Corded Ware peoples of central Europe5. However, the fact that the Srubnaya also harbored such ancestry indicates that the Anatolian Neolithic or EEF ancestry could have come into the steppe from a more eastern source."

But IMO this is not sufficient; as Sample I0432 also had such ancestry it should cause a complete rethinking of autosomal relationships of Corded Ware vis a vis Sintashta and Andronovo.

Srubnaya postdates Sintashta and is mixed with local Yamnaya /Poltavka subcultures so I don't get how EEF among them is disproving the Sintashta -Corded Ware connection, when Sintashta is much older and had more EEF. This statement of Matthieson et al. is naive and they ignored that EEF was not present south and east of Eastern Yamnaya, where the EEF-admixed ancestors of Sintashta moved later in.

Coldmountains
11-30-2015, 08:19 PM
But I was actually expecting R1a-Z93 to be a dominant haplogroup in Eastern Yamna and in the descending Poltavka culture, while it turned out that Z93 dominated the steppe much later and was initially a part of the broad Corded Ware Horizon (or its Eastern part that produced the Abashevo culture, ancestral to Sintashta), or at least this is what is quite strongly suggested by the autosomal results of the Sintashta samples. This association with Corded Ware and Abashevo was something I considered a possible but, nevertheless, less likely scenario (http://www.anthrogenica.com/showthread.php?1516-R1a-and-Corded-Ware&p=17642&viewfull=1#post17642).

When I first heard that Yamnayas were rather "dark" pigmented and had a lower frequency of light eyed people than modern European populations I became skeptical about the association of Yamnaya with Indo-Iranians because we knew already quite early that Proto-Indo-Iranians and steppe Iranians were quite light pigmented and europid looking but I still expected them to be mainly R1a but I was totally wrong. I also thought that Proto-Indo-Iranians were very light but had much of that "teal" admixture even more than Yamnaya but I was partially wrong here because Sintashta and Andronovo had much less of it than Yamnaya. I just assumed that Proto-Indo-Iranians must have much of it because most modern Indo-Iranians have very much of it especially the Indo-Iranians richest in R1a.

Michał
11-30-2015, 08:33 PM
Srubnaya postdates Sintashta and is mixed with local Yamnaya /Poltavka subcultures so I don't get how EEF among them is disproving the Sintashta -Corded Ware connection, when Sintashta is much older and had more EEF. This statement of Matthieson et al. is naive and they ignored that EEF was not present south and east of Eastern Yamnaya, where the EEF-admixed ancestors of Sintashta moved later in.
It seems that the wording used by Matthieson was very unfortunate, as it may lead to two completely different interpretations, namely that the source of Sintashta's/Srubna's EEF was located 1) east of Central Europe (for example in Southern Belarus or NW Ukraine, which makes perfect sense) or 2) east of the Volga steppe or even east of the Ural Mountains (which doesn't make any sense at all).

parasar
11-30-2015, 09:14 PM
Srubnaya postdates Sintashta and is mixed with local Yamnaya /Poltavka subcultures so I don't get how EEF among them is disproving the Sintashta -Corded Ware connection, when Sintashta is much older and had more EEF. This statement of Matthieson et al. is naive and they ignored that EEF was not present south and east of Eastern Yamnaya, where the EEF-admixed ancestors of Sintashta moved in.

We don't know that as yet.

We do know that Sample I0432 which as old as if not older than CW samples, had it.

I think we have to first determine the Eastern spread of EEF.
Was it present in Iraq, Iran, for example? "Cranial analysis ... (BMAC) was formed from a population migrating from north-west Iran" The Origin of the Indo-Iranians By Elena Efimovna Kuzʹmina

Why is there this Y-DNA bifurcation - see eg.
"Punjabi G-L166 sample, HG02681, here shares four L166-specific SNPs with the 5,300-year-old G-L166 Iceman mummy Ötzi, found in the Italian part of the Ötztal Alps [31] (data from ERP001144), and the same number with a subset of G-L91 samples from Corsica, Sardinia or Tuscany, inferred to be also G-L166"
"J2a-PF4610* is revealed by a Punjabi sample and independently confirmed by seven Sardinian SNP results"
"J2a-PF5197/Z2397 is a new haplogroup with 13 samples belonging to two subclades. The most important is J2aPF5174,
with three sequences from Tamil, two Bengali and two Punjabi as well as one Gujarati. When 13 Sardinian
samples are also considered, a distribution from South Asia to Europe is apparent."

We also have potentially J1, J2b2, H, and F lines that spread in the Neolithic. There are many more dots to connect.

Jean M
11-30-2015, 10:05 PM
I was arguing on Rootsweb, at FTDNA's forum, and on the old dna forums for that and the connection of both R1b and R1a to the early Indo-Europeans as long ago as 2006. Of course, I am neither famous nor a professional scientist, so I was mostly laughed off.

Not by me. My online Peopling of Europe followed you on this point from its first version in March 2009. I'm sure about that, though I never kept a copy. There seemed no point, as it was constantly in the process of revision. It might be amusing to see how many errors I made then. But following your thinking was not one of them. :)

Jean M
11-30-2015, 10:12 PM
Nevertheless we see I1 shared SNPs in specimens both from Hungarian LBK (http://biorxiv.org/content/biorxiv/early/2014/09/03/008664.full.pdf) and Mesolithic Sweden (Stora Forvar 11) .....

The claim of I1 in Stora Förvar 11 is by Genetiker, not Skoglund 2014, and needs to be treated with caution. See comments http://www.anthrogenica.com/showthread.php?5279-Misc-I1-and-I2-stuff-from-R1b-threads

yxc
12-01-2015, 01:29 AM
treated with caution should had been any conclusions that were deliberately made when the autosomals of up to 3500bc Gokhem Swedish farmers were published
No explaination was given on how and why the Oxygen isotope analysis of their Enamel determined where they grew up.
As Colonists generations ago they had came to Scania. Their autosomal showed that they were like midlle Elbe farmers from whom they obviously descent from .

The so called Steppe ancestry had just been certain mtdna that weren't in the farmer settlements but all over the east-eur. mesolthic.
If Gotland had these mtdna then vast Scandinavia had them too + uncertain Ydna just like had perhaps Ostorf ,mesolithic Zvejnieky , Dabki all the way to Elbe and who knows if not to the Seine and the Isles