PDA

View Full Version : Birthplace of Y-DNA DE and E - Africa or Eurasia?



Lank
07-14-2015, 05:12 PM
The new paper by Trombetta et al. (http://gbe.oxfordjournals.org/content/7/7/1940) reassigned 3 East Africans carrying E* to a novel lineage, of distant common ancestry with E-P147 (current "E1"). Here's (http://i61.tinypic.com/rifbiw.png) the updated phylogeny. This also got me thinking about the origin of E and DE.

For those who aren't aware, there has been a long-standing debate on whether DE is African or Asian in origin. With the increasing prevalence of Y chromosome sequencing, since recently there has been improved dating of Y-DNA lineages, courtesy of YFull (http://www.yfull.com/tree/CT/).

YFull estimates proto-Eurasian CT to have bifurcated 68 kya. The TMRCA of DE and CF are slightly younger, dated to ~66 kya. If we look at these founding fathers, there are some interesting patterns. The definitively Eurasian D/C/F lineages do not break apart until 48 kya, when all three undergo rapid diversification. To use an example, a mere <2 ky separate the TMRCA of F (~48 kya) from the formation of pre-IJ.

This would be a time when proto-Eurasians (similar to Ust'-Ishim (http://www.nature.com/nature/journal/v514/n7523/full/nature13810.html)?) were expanding in various directions. The fact that D/C/F don't seem to break apart into daughter lineages for nearly 20,000 years following the origin of DE/CF, most likely corresponds with the population bottleneck, which is also visible in the autosomal DNA of modern Eurasians.

The story of E is quite different. 54 kya, the ancestors of E1 and E2 had already begun to form, and 53 kya, the ancestors of E1a and E1b had already begun to split from one another. I would say these patterns support that both CF and DE made it out of Africa, where only DE survived (DE is very close in age to CT anyway), and E would later diversify.

Those who advocate an Asian origin of Y-DNA DE or E must explain why E exhibits a distinct evolutionary history from the clearly Eurasia-derived lineages. The only way out, as I see it, is to suggest that E originated and diversified in Asia, but this would create even more problems. Where in Eurasia could E diversify so early on, while the ancestors of C/D/F underwent a bottleneck in some other location, and how was E wiped out from later Eurasians?

I would be interested in hearing some other perspectives on this.

Kale
07-15-2015, 02:09 PM
Well the parallels of history E shares the with D,C, and F leads me to believe they were never too far from each other until pretty late.
100kya: DT splits from B.
73kya: DE and CT split.
70kya: D splits from E, C splits from F.
55kya: E splits into E1 and E2.
53kya: Pre-G and pre-H break off of F.
52kya: The earliest split in C occurs.
50kya: The earliest split in D occurs.

2,000 years is not a lot in this context.

Whatever location C,D,F diversified in is most likely the same place E diversified in, wherever that may be.

EDIT: Also, never underestimate lineage replacement.
Oase: K in Romania 40kya. No basal K there today.
K14: C1(b?) in Russia 37kya. Where is that now?
MA1: R* in Siberia 24kya. There isn't any there now.
Mesolithic Europe: I believe C1a2 was something like 3/13 specimens sampled so far. Where is that now?
Mesolithic Scandinavia: All I2a. Now I2a is less than 1% in areas right on top of where the ancient sampled were recovered.

Lank
07-15-2015, 03:57 PM
Where are those estimates coming from? I was looking at YFull, where the earliest splits within C, F, and D, are all dated to 48 kya. I found it rather remarkable that these Eurasian founding fathers had sons branching off at the same time, unlike African-affiliated E, which bifurcated ~6000 years earlier.

Also, we do find DE(xE) in Africa, and it is found in West Africa, rather than the north/east which has seen evident back-migrations from Eurasia. DE is very close in age to CT, so they should have originated in essentially the same population.

I would agree that the early ancestral nodes of C, F, D, and E most likely originated in the same population. However, IMO there is no reason at present to assume that the descendents of C/F/D/E would have branched off in the same population. Firstly, as previously mentioned, E splits off slightly earlier than the other three. Secondly, from archaeology, we can infer that these lineages branched off during a period when vast areas of the world were being colonized, in the early stages of the Late Stone Age and Upper Paleolithic technological revolutions. So there was a lot of population movement.

I am aware that modern distributions can be deceiving. However, it is possible to work with what we have in order to judge probabilities, based on the present evidence. Considering the presence of DE* in Africa, which is close in age not only to its CT ancestor, but also to mtDNA L3, and the deviant evolutionary history of E (if YFull is to be trusted), I believe an African origin is the most parsimonious scenario. It is not difficult to imagine that DE, which originated shortly after CT, may have originated in Africa just prior to OOA, while its CF sibling, which was still in its very earliest stages, could easily have gone extinct in Africa. I should add that mtDNA L3 is the likely female counterpart of CT, is of a similar age, and appears even more clearly African (East African to be more specific) based on the modern distribution.

Of course, as always, ancient DNA is the only way to get a definitive answer. I am just somewhat puzzled by the fact that some can be so confident in a Eurasian origin of Y-DNA E, when a strong case exists for an African origin, although it could go either way in the end since this was perhaps our most mobile period as a species.

Kale
07-16-2015, 01:04 PM
Source: A recent bottleneck of Y chromosome diversity coincides with a global change in culture

We do find DE in West Africa yes, but we also find DE*(xD,xE) in Tibet. That means, that DE has been effectively wiped out between those two locations. Never underestimate lineage replacement. The divergence between CT and DE roughly coincides with the divergence between Mtdna M and N. Y-Dna D does not occur in populations without huge amounts of mtdna M. Populations like Andaman and Onge are fixed for both if I remember correctly. Likewise in West Eurasia, beyond the small trickle of E that came with the neolithic, the people are 100% y-dna CF, and 100% mtdna N.

I personally think that the origins of C,D,E,F are all intertwined up until 50,000 years ago. I also think an African origin of E is likely, more specifically Northeast Africa.

EDIT: We do have a tiny bit of ancient DNA from Africa. There was a study regarding Sudanese remains from mesolithic to the post-christian era. Though the sample size was small, the mesolithic and neolithic samples were all ydna A or B. E did not show up until the late/post-neolithic. F not showing up until christian times.

malleus
07-16-2015, 01:43 PM
Oase: K in Romania 40kya. No basal K there today.

Oase is conservatively F*. probably IJK*
If one is to accept Genetiker's result and speculatively assign him to K then it only makes sense to assign him to K2a as well.
Not that this goes against what you are claiming but just to clarify the details.

lgmayka
07-16-2015, 01:55 PM
We do find DE in West Africa yes, but we also find DE*(xD,xE) in Tibet.
DE(xD,xE) has also been found in Syria. See FTDNA kit 130501, in the Early Branches project (https://www.familytreedna.com/public/Y-DNA-Tree-Early-Branches/default.aspx?section=yresults). Is anyone willing to offer to pay for his Big Y? :)

Lank
07-16-2015, 05:34 PM
Source: A recent bottleneck of Y chromosome diversity coincides with a global change in culture
I checked that source, and what I found was a date for GT of 51,500 years ago (all other splits in C/F/D are younger), compared to just over 54,000 years ago for E. I don't think these results are very different from YFull's estimates.


We do find DE in West Africa yes, but we also find DE*(xD,xE) in Tibet. That means, that DE has been effectively wiped out between those two locations. Never underestimate lineage replacement. The divergence between CT and DE roughly coincides with the divergence between Mtdna M and N. Y-Dna D does not occur in populations without huge amounts of mtdna M. Populations like Andaman and Onge are fixed for both if I remember correctly. Likewise in West Eurasia, beyond the small trickle of E that came with the neolithic, the people are 100% y-dna CF, and 100% mtdna N.

I personally think that the origins of C,D,E,F are all intertwined up until 50,000 years ago. I also think an African origin of E is likely, more specifically Northeast Africa.
Thanks for the response. A possible link between Y-DNA D and mtDNA M is very interesting, but it might be an artifact of generally lower Y-DNA/mtDNA diversity in modern West Eurasians, compared to East Eurasians. Things may have been different in the past. Pre-Neolithic Europeans do seem to have had fairly high frequencies of Y-DNA C, whereas modern Europeans tend to carry F-derived lineages (excluding E).


EDIT: We do have a tiny bit of ancient DNA from Africa. There was a study regarding Sudanese remains from mesolithic to the post-christian era. Though the sample size was small, the mesolithic and neolithic samples were all ydna A or B. E did not show up until the late/post-neolithic. F not showing up until christian times.
I remember that study. It's certainly an interesting find. Even today, the E diversity and frequency of Sudanese is low compared to their northern, eastern and western neighbors. There is not a single E lineage that would appear to have originated in Sudan based on modern distributions.


DE(xD,xE) has also been found in Syria. See FTDNA kit 130501, in the Early Branches project (https://www.familytreedna.com/public/Y-DNA-Tree-Early-Branches/default.aspx?section=yresults). Is anyone willing to offer to pay for his Big Y? :)

I'd say all of the DE* samples should be sequenced. Sequencing that sample alone would reveal very little, if we don't understand how it relates to West African and Tibetan DE*. For the record, recent African lineages (not saying the Syrian DE* necessarily is) are usually found in MENA populations at low frequencies.

We also see a signature in the autosomal DNA of Lebanese (http://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1003316) and Egyptian (http://www.nature.com/articles/srep09996) Muslims of recent low-level West African admixture lacking in the Christian minorities. The minor African ancestry in the Copts and Lebanese Christians is strictly East African, unlike in the Muslims. This is important since DE* has been found in West Africa, not East Africa.

Passa
07-17-2015, 06:51 PM
......

Kurumim
07-11-2017, 04:35 AM
DE(xD,xE) has also been found in Syria. See FTDNA kit 130501, in the Early Branches project (https://www.familytreedna.com/public/Y-DNA-Tree-Early-Branches/default.aspx?section=yresults). Is anyone willing to offer to pay for his Big Y? :)
woah! thats amazing

Chad Rohlfsen
07-16-2017, 10:53 AM
Source: A recent bottleneck of Y chromosome diversity coincides with a global change in culture

We do find DE in West Africa yes, but we also find DE*(xD,xE) in Tibet. That means, that DE has been effectively wiped out between those two locations. Never underestimate lineage replacement. The divergence between CT and DE roughly coincides with the divergence between Mtdna M and N. Y-Dna D does not occur in populations without huge amounts of mtdna M. Populations like Andaman and Onge are fixed for both if I remember correctly. Likewise in West Eurasia, beyond the small trickle of E that came with the neolithic, the people are 100% y-dna CF, and 100% mtdna N.

I personally think that the origins of C,D,E,F are all intertwined up until 50,000 years ago. I also think an African origin of E is likely, more specifically Northeast Africa.

EDIT: We do have a tiny bit of ancient DNA from Africa. There was a study regarding Sudanese remains from mesolithic to the post-christian era. Though the sample size was small, the mesolithic and neolithic samples were all ydna A or B. E did not show up until the late/post-neolithic. F not showing up until christian times.

What DE in Africa? You aren't talking about those 15-20 year old studies based on a couple STRs are you? I'm not aware of anyone coming back positive in Africa.

Passa
07-16-2017, 01:03 PM
What DE in Africa? You aren't talking about those 15-20 year old studies based on a couple STRs are you? I'm not aware of anyone coming back positive in Africa.

DE* cases have been found in West Africa (especially SE Nigeria) in SNP-based studies. Personally I think that what we are seeing are back-mutations though.

Chad Rohlfsen
07-16-2017, 03:18 PM
DE* cases have been found in West Africa (especially SE Nigeria) in SNP-based studies. Personally I think that what we are seeing are back-mutations though.

Can you provide the study that has these? Back mutations would make more sense. People trying to divorce E from ancestral Eurasians is ridiculous. I'm not saying you're doing that though. Every African is Eurasian admixed. Time and ancient DNA will show this. I have no doubt. The San are down. Next is the Yoruba and Mota. We're all mixed. No one is pure anything, other than pure Eurasian, maybe.

Megalophias
07-16-2017, 03:33 PM
Can you provide the study that has these? Back mutations would make more sense.

The original study (http://www.genetics.org/content/165/1/229.full) tested for 3 different E mutations and 1 D and they were all negative. I don't know what makes you so confident that E must be Eurasian.

Chad Rohlfsen
07-16-2017, 03:53 PM
CT is Eurasian. CF and D are too. Why would E be a special case? The split between Eurasians and SSA is likely older than CT. No one really disputes that. It takes a lot of special pleading to separate one lineage from Eurasians, but not a single other.

You're not mistaking my ancestral Eurasians comment as meaning anything other than non-SSA, right? Also, that study is from 2003. Exactly what I warn against. Show me a recently tested DE in SSA.

Passa
07-16-2017, 03:59 PM
Can you provide the study that has these? Back mutations would make more sense. People trying to divorce E from ancestral Eurasians is ridiculous. I'm not saying you're doing that though. Every African is Eurasian admixed. Time and ancient DNA will show this. I have no doubt. The San are down. Next is the Yoruba and Mota. We're all mixed. No one is pure anything, other than pure Eurasian, maybe.

Indeed that's what I think too.

This study found several samples belonging to DE(xE) in Cross River state, Nigeria - https://bmcevolbiol.biomedcentral.com/articles/10.1186/1471-2148-10-92

Chad Rohlfsen
07-16-2017, 04:37 PM
Indeed that's what I think too.

This study found several samples belonging to DE(xE) in Cross River state, Nigeria - https://bmcevolbiol.biomedcentral.com/articles/10.1186/1471-2148-10-92

Yeah, another old one. Re-testing would likely change those outcomes. DNA studies even from 2009 to today is like comparing the Stone Age to the Industrial Revolution.

MfA
07-16-2017, 04:46 PM
Yeah, another old one. Re-testing would likely change those outcomes. DNA studies even from 2009 to today is like comparing the Stone Age to the Industrial Revolution.

There's an African-American guy on the 23andMe forums who's DE.

Megalophias
07-16-2017, 05:14 PM
CT is Eurasian. CF and D are too. Why would E be a special case? The split between Eurasians and SSA is likely older than CT. No one really disputes that. It takes a lot of special pleading to separate one lineage from Eurasians, but not a single other.
This is a small number of lineages so how many occur in different continents can be a result of chance. Looking at CDEF's female counterpart L3 we have 5 African and 2 Eurasian branches.

Of course it will be associated with "ancestral Eurasians" of some kind, whether those who actually left Africa or a sister group that remained in Africa. SSAs have a mix of autosomal ancestry of various degrees of relationship to Eurasians. Sub-Saharan Africa one way or another has a variety of basal E and L3, so why dismiss DE*? What difference does it make whether it came with Eurasians or not?


Also, that study is from 2003. Exactly what I warn against. Show me a recently tested DE in SSA.
Repeatedly checking multiple equivalent SNPs is not weak because it was done in 2003. Found again in Guinea-Bissau in 2007. And 2010. Who is special pleading here exactly?

Lupriac
11-10-2018, 08:27 PM
[Deleted]