PDA

View Full Version : E-M84 - A much-needed update and a call for collaboration



leorcooper19
03-24-2021, 05:43 PM
Hello everyone,

Here, I'll gather some of the newer updates and meaningful additions to E-M84's story. This macrogroup is very important to the population history of SW Asia and all surrounding regions, and hasn't gotten nearly enough focus.

To start, I'd like to share that a scientific sample (labeled as Uzbekistan_Jew)- not yet on YFull- from this study (https://www.biorxiv.org/content/10.1101/2020.10.18.342816v1) has been found to split E-M84, being negative for 20 SNPs that are currently at the E-M84 level on YFull. This sample is on FTDNA, as is responsible for the difference between E-L795 (phyloequivalent with M84) and E-S11835 (the new name for the level shared by everyone except for the Uzbekistan_Jew). The new TMRCA of E-M84 with this sample would be around 10,000 ybp. I'm trying to get the sample onto YFull but will need to convert the .tsv.bz2 file into a BAM file first.

Summing up the phylogeny as it exists now, E-M84 has several large groups that can be spoken about in a unified way as well as many more basal/isolated lines that are all pretty unique. (I'll be using the v8.09 Tree from YFull as the latest trees to not have TMRCAs for E)
1) To me, it looks like E-Y5435 (https://www.yfull.com/arch-8.09/tree/E-Y5435/) and E-PF6751 (https://www.yfull.com/arch-8.09/tree/E-PF6751/) can be plausibly linked with certain subclades of J-P58 and T-L208 that are associated with the spread of the Semitic family throughout the region. From subject experts of those two groups, I'd love to know which specific subgroups jump out as analogies and anything about their histories and spread after their TMRCAs. A few steps downstream E-PF6751 at E-CTS9608 (https://www.yfull.com/arch-8.09/tree/E-CTS9608/) we have the oldest E-M84 sample to date, coming in from Arslantepe c. 5200 ybp. This, combined with basal samples from Lebanon in both J1-FGC11 and J2b-M205, to me shows that the spread of major Semitic-associated Y-DNA lineages happened on a north-to-south axis.

2) In terms of the more basal/isolated lineages, I see:
- the unmarked E-M84* (https://www.yfull.com/arch-8.09/tree/E-M84*/) (probably in E-B406, if so then they'd be a Gulf Arab)
- E-Y82779 (https://www.yfull.com/arch-8.09/tree/E-Y82779/), which is a peculiar lineage with a basal Syrian Jew and a relatively young European lineage that isn't made up by Ashkenazi Jews.
- the basal Spaniards at E-S11387* (https://www.yfull.com/arch-8.09/tree/E-S11387*/) and E-CTS5265* (https://www.yfull.com/arch-8.09/tree/E-CTS5265*/)
- E-Y60961 (https://www.yfull.com/arch-8.09/tree/E-Y60961/), which is an old subclade but is only found in Iraqi, Syrian, and Sephardic Jews, as well as Palestinians. Probably is a deeply rooted Levantine clade.
- My own E-Y14899 (https://www.yfull.com/arch-8.09/tree/E-Y14899/), which contains three relatively young clades made up by Ashkenazi Jews in one, Gulf Arabs in one, and Palestinian Christians and a Cypriot in another.

All-in-all, it seems quite likely that the MRCA of E-M84 and E-S11835 lived in the Levant, with the big groups E-Y5435 and E-PF6751 finding significant success throughout SW Asia and beyond, while the more basal lineages stayed in the Levant for much longer and left much later, and thus couldn't grow as much as their bigger cousins.

Next, I'll list some of the major questions that are still unanswered:
1) Does E-M84 directly descend from Natufian (or a related group's) E-Z830, or a more recent migration from Northeast Africa? In other words, what is the origin of E-M34? E-Z841's spread may imply deep routes outside of NE Africa, but the presence of E-M34 in many NE Africans makes me think that we are underestimating the region as a source of unique E-M35 lines.
2) Does E-M84 really originate in the Levant? While I think it's hard to say for sure, why couldn't have E-M84 originated in the upper fertile crescent? Samples such as RISE423 from Lake Sevan, Armenia c. 1400-1200 BCE who are E-M84 (xY5412, xY14899, xPF7651, credit to MfA here (https://anthrogenica.com/showthread.php?8672-List-of-ancient-samples-on-GEDmatch&p=192146&viewfull=1#post192146)) may imply that the Levant was a kind of sink for E-M84 lineages, similar to R-M269 in Western Europe, and that we just need the relevant samples from the right region to know.
3) What other lineage is the best analogy to E-M84? After looking around on YFull, I have to say that T-L208 overall resembles E-M84 in many important ways. Does anyone else have any other suggestions?

Please share any information not covered in this post here on the thread. I appreciate it and look forward to some much-needed discussion.

capsian
03-24-2021, 06:32 PM
Hello everyone,

Here, I'll gather some of the newer updates and meaningful additions to E-M84's story. This macrogroup is very important to the population history of SW Asia and all surrounding regions, and hasn't gotten nearly enough focus.

To start, I'd like to share that a scientific sample (labeled as Uzbekistan_Jew)- not yet on YFull- from this study (https://www.biorxiv.org/content/10.1101/2020.10.18.342816v1) has been found to split E-M84, being negative for 20 SNPs that are currently at the E-M84 level on YFull. This sample is on FTDNA, as is responsible for the difference between E-L795 (phyloequivalent with M84) and E-S11835 (the new name for the level shared by everyone except for the Uzbekistan_Jew). The new TMRCA of E-M84 with this sample would be around 10,000 ybp. I'm trying to get the sample onto YFull but will need to convert the .tsv.bz2 file into a BAM file first.

Summing up the phylogeny as it exists now, E-M84 has several large groups that can be spoken about in a unified way as well as many more basal/isolated lines that are all pretty unique. (I'll be using the v8.09 Tree from YFull as the latest trees to not have TMRCAs for E)
1) To me, it looks like E-Y5435 (https://www.yfull.com/arch-8.09/tree/E-Y5435/) and E-PF6751 (https://www.yfull.com/arch-8.09/tree/E-PF6751/) can be plausibly linked with certain subclades of J-P58 and T-L208 that are associated with the spread of the Semitic family throughout the region. From subject experts of those two groups, I'd love to know which specific subgroups jump out as analogies and anything about their histories and spread after their TMRCAs. A few steps downstream E-PF6751 at E-CTS9608 (https://www.yfull.com/arch-8.09/tree/E-CTS9608/) we have the oldest E-M84 sample to date, coming in from Arslantepe c. 5200 ybp. This, combined with basal samples from Lebanon in both J1-FGC11 and J2b-M205, to me shows that the spread of major Semitic-associated Y-DNA lineages happened on a north-to-south axis.

2) In terms of the more basal/isolated lineages, I see:
- the unmarked E-M84* (https://www.yfull.com/arch-8.09/tree/E-M84*/) (probably in E-B406, if so then they'd be a Gulf Arab)
- E-Y82779 (https://www.yfull.com/arch-8.09/tree/E-Y82779/), which is a peculiar lineage with a basal Syrian Jew and a relatively young European lineage that isn't made up by Ashkenazi Jews.
- the basal Spaniards at E-S11387* (https://www.yfull.com/arch-8.09/tree/E-S11387*/) and E-CTS5265* (https://www.yfull.com/arch-8.09/tree/E-CTS5265*/)
- E-Y60961 (https://www.yfull.com/arch-8.09/tree/E-Y60961/), which is an old subclade but is only found in Iraqi, Syrian, and Sephardic Jews, as well as Palestinians. Probably is a deeply rooted Levantine clade.
- My own E-Y14899 (https://www.yfull.com/arch-8.09/tree/E-Y14899/), which contains three relatively young clades made up by Ashkenazi Jews in one, Gulf Arabs in one, and Palestinian Christians and a Cypriot in another.

All-in-all, it seems quite likely that the MRCA of E-M84 and E-S11835 lived in the Levant, with the big groups E-Y5435 and E-PF6751 finding significant success throughout SW Asia and beyond, while the more basal lineages stayed in the Levant for much longer and left much later, and thus couldn't grow as much as their bigger cousins.

Next, I'll list some of the major questions that are still unanswered:
1) Does E-M84 directly descend from Natufian (or a related group's) E-Z830, or a more recent migration from Northeast Africa? In other words, what is the origin of E-M34? E-Z841's spread may imply deep routes outside of NE Africa, but the presence of E-M34 in many NE Africans makes me think that we are underestimating the region as a source of unique E-M35 lines.
2) Does E-M84 really originate in the Levant? While I think it's hard to say for sure, why couldn't have E-M84 originated in the upper fertile crescent? Samples such as RISE423 from Lake Sevan, Armenia c. 1400-1200 BCE who are E-M84 (xY5412, xY14899, xPF7651, credit to MfA here (https://anthrogenica.com/showthread.php?8672-List-of-ancient-samples-on-GEDmatch&p=192146&viewfull=1#post192146)) may imply that the Levant was a kind of sink for E-M84 lineages, similar to R-M269 in Western Europe, and that we just need the relevant samples from the right region to know.
3) What other lineage is the best analogy to E-M84? After looking around on YFull, I have to say that T-L208 overall resembles E-M84 in many important ways. Does anyone else have any other suggestions?

Please share any information not covered in this post here on the thread. I appreciate it and look forward to some much-needed discussion.
I believe that the Haplogroup E-Z830 originated from the Sinai Peninsula and migrated more than 19000 years ago to the Levant

Agamemnon
03-24-2021, 07:13 PM
In terms of parallels between J1-P58 and E-M84, here are a few ones which stand out:


E-Y60961 looks quite similar in terms of TMRCA and overall distribution to J-Y15152 (https://yfull.com/tree/J-Y15152/).


Several of the major sub-branches of E-Y5435 and E-PF6751 have TMRCA estimates that fall between 4600 & 4000 yBP, there's a parallel here with J-Z1884 (https://yfull.com/tree/J-Z1884/) and its many sub-branches (FGC12834 (https://yfull.com/tree/J-FGC12834/), Z640 (https://yfull.com/tree/J-Z640/), Y2919 (https://yfull.com/tree/J-Y2919/)) as well as other major branches like J-FGC4745 (https://yfull.com/tree/J-FGC4745/). This, in all likeliness, corresponds with the dissemination and break-up of Central Semitic.


Under Y5435, Y44427 resembles J-FGC3723 (https://yfull.com/tree/J-Y6370/), which makes it a prime contender for an association with Epigraphic South Arabian.


Y18691 is strikingly similar to J-S21237 (https://yfull.com/tree/J-S21237/), the two probably expanded together with the earliest Arab dispersals.


This is a non-exhaustive list, I'm sure there are many more parallels of the sort. A post-Natufian arrival for E-M34 makes the most sense in my view, as you said its presence in other AA-speaking groups in NE Africa is too often overlooked, M84 on the other hand easily could have originated in-situ in the Levant. Odds are that this lineage (along with L791) is tied to the Afroasiatic-speaking community that brought the Semitic Parent Language to the Levant from NE Africa, the phylogenetic relationship to E-L19 via Z827 does seem to reflect the Semito-Berber node though the correlation isn't exactly perfect (age being the major factor). Z830 possibly made it to the Levant by the Kebaran period.

Helves
03-24-2021, 09:13 PM
You're asking some important questions about E-M34 and M84. It's interesting to me that although M34 has a TMRCA that matches proto-AA and was in all likelihood still in Egypt/NE Africa thousands of years after it's origin, the distribution M34(xM84) has today looks undoubtedly more Near Eastern than anything with few if any basal North African branches.

drobbah
03-24-2021, 09:27 PM
the phylogenetic relationship to E-L19 via Z827 does seem to reflect the Semito-Berber node though the correlation isn't exactly perfect (age being the major factor). Z830 possibly made it to the Levant by the Kebaran period.
Considering you stated before that Cushitic also belonged to this node but branched off earlier from the Semito-Berbers, is it plausible that E-Z830 could represent the early Cushitic-Semitic-Berber community as the PNs and their modern SE Africans descendants are dominated by E-M293, other E-Z830 lineages survived to this day in the Horn like E-V6 and others.Also do you have any ideas on how E-M78 fits into all of this and AA dispersals, it is a lineage that is very important among modern Afro-Asiatic speakers and there's also the case of a E-M78+ Neolithic Levantine remain.

Agamemnon
03-25-2021, 07:04 PM
Considering you stated before that Cushitic also belonged to this node but branched off earlier from the Semito-Berbers, is it plausible that E-Z830 could represent the early Cushitic-Semitic-Berber community as the PNs and their modern SE Africans descendants are dominated by E-M293, other E-Z830 lineages survived to this day in the Horn like E-V6 and others.Also do you have any ideas on how E-M78 fits into all of this and AA dispersals, it is a lineage that is very important among modern Afro-Asiatic speakers and there's also the case of a E-M78+ Neolithic Levantine remain.

This certainly is a possibility, M78 however is an equally good fit for the Cushitic-Semito-Berber node. We have V32, the main Cushitic marker, and V65, which tracks the Libyan branch of Libyco-Berber, under Z1902 (https://www.yfull.com/arch-8.09/tree/E-Z1902/). In some respects, it seems to be a better fit for Afroasiatic, M78's TMRCA seemingly correlates with the dissolution of Common Afroasiatic, this has to mean something because we also have correlates for the Egyptian branch in the form of V22 and V12 (xV32) the latter also being under Z1902. While there are correlates for the Omotic, Semito-Berber and Chadic branches, the best ones are either under Z830 or lineages such as J-P58, J-M205, T-L208 or R-V88 (which correlate with discrete branches, the last two might correlate with several branches though indicating an early introgression).

I think the takeaway here is that there is no perfect correlation, and that we should not expect one. The earliest AA groups probably carried all the branches of M35 we've discussed here. Keep in mind that PAA as we currently think of it could be a collection of closely related dialects, and that some of the branches essential to reconstructing the proto-language are either too poorly documented (Omotic, Chadic) or unknown to us, having disappeared and/or been replaced by some other branch. As in the natural realm, extinction is the norm while survival is the exception in diachronic linguistics, most languages spoken by man are now long-gone. It's the same story at a more granular level, the correlation between R1 lineages and the IE language tree (which is the level of accuracy we should aim for) is far from perfect, some branches align with R1a and others with R1b, rinse and repeat with Semitic (J-P58 is going to be predominant in some branches, M34 in others, etc).

capsian
03-25-2021, 07:50 PM
In terms of parallels between J1-P58 and E-M84, here are a few ones which stand out:


E-Y60961 looks quite similar in terms of TMRCA and overall distribution to J-Y15152 (https://yfull.com/tree/J-Y15152/).


Several of the major sub-branches of E-Y5435 and E-PF6751 have TMRCA estimates that fall between 4600 & 4000 yBP, there's a parallel here with J-Z1884 (https://yfull.com/tree/J-Z1884/) and its many sub-branches (FGC12834 (https://yfull.com/tree/J-FGC12834/), Z640 (https://yfull.com/tree/J-Z640/), Y2919 (https://yfull.com/tree/J-Y2919/)) as well as other major branches like J-FGC4745 (https://yfull.com/tree/J-FGC4745/). This, in all likeliness, corresponds with the dissemination and break-up of Central Semitic.


Under Y5435, Y44427 resembles J-FGC3723 (https://yfull.com/tree/J-Y6370/), which makes it a prime contender for an association with Epigraphic South Arabian.


Y18691 is strikingly similar to J-S21237 (https://yfull.com/tree/J-S21237/), the two probably expanded together with the earliest Arab dispersals.


This is a non-exhaustive list, I'm sure there are many more parallels of the sort. A post-Natufian arrival for E-M34 makes the most sense in my view, as you said its presence in other AA-speaking groups in NE Africa is too often overlooked, M84 on the other hand easily could have originated in-situ in the Levant. Odds are that this lineage (along with L791) is tied to the Afroasiatic-speaking community that brought the Semitic Parent Language to the Levant from NE Africa, the phylogenetic relationship to E-L19 via Z827 does seem to reflect the Semito-Berber node though the correlation isn't exactly perfect (age being the major factor). Z830 possibly made it to the Levant by the Kebaran period.

so E-M123 we can said is Branch origin Levant though ancestor come from North east africa is entered the Levant 20,000 yrs ago and this is very old

leorcooper19
03-25-2021, 09:26 PM
You're asking some important questions about E-M34 and M84. It's interesting to me that although M34 has a TMRCA that matches proto-AA and was in all likelihood still in Egypt/NE Africa thousands of years after it's origin, the distribution M34(xM84) has today looks undoubtedly more Near Eastern than anything with few if any basal North African branches.

It does indeed look like that on FTDNA and YFull, but Cruciani et al. 2004 found that E-M34 is relatively common among several Ethiopian groups (8/34 = 23.5% in Amharas, 3/22 = 13.6% in Ethiopian Jews, 1/12 = 8.3% in Wolaytas, and 2/25 = 8% in Oromos). Are we to believe all of these are from E-M84 via Arabia? Or could there be a missing, third branch of E-M34 deep in Africa?

Indeed, Trombetta et al. 2011 (https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0016073#pone.0016073.s001) found that "the V23 mutation was found to mark a new lineage within the E-M34 clade. Two out of 16 E-M34 Y chromosomes which had been previously observed in Africa [8] turned out to carry this mutation." The samples seem to originally be from Semino et al 2004, and they found that E-M123 exists at rates of 4/78 = 5.1% in Oromos and 1/48 = 2.1% in Amharas. In fact, these are the only Sub-Saharan Africans who have any E-M123, so one (or both) of these Ethiopian groups must be the source of the E-V23 clade that is very possibly a third subclade of E-M34, located far from its cousins and deep in what could be the real homeland of E-M34.

Anyone have any idea if any other Ethiopian E-M34 sequences can be found throughout the literature or if any samples can be found on a site like FTDNA? I for one would be happy to pay for a test that officially puts this clade on the tree.

drobbah
03-25-2021, 10:06 PM
The Omotic groups have high E-M34, if I recall correctly from Awale's blog there's an Omotic ethnicity that carries E-M34 at 25%.Perhaps the earliest Omotic speakers were responsible for bringing it to the Horn from NE Africa

drobbah
03-26-2021, 09:32 AM
The Hirbo thesis found:
9% of E-M123 in Ethiopian_Burji (Highland East Cushitic, Southern Ethiopia)
8% of E-M123 in Kenyan_Yaku (Endangered Cushitic speakers, unknown position in the family)
5% of E-M123 in Kenyan_Boni (Somali-related language speakers who are hunter-gatherers)

Keneki20
03-28-2021, 05:04 PM
Can anyone hazard an educated guess as to any archaeological/cultural correlate(s) that brought E-Z830 to the Levant in the Epipaleolithic? The Mushabian culture as the correlate is oft-stated, but has lots of problems. The notion that the Mushabian culture has roots in North Africa doesn’t seem to be so tenable anymore, as most recent publications I’ve looked at and archaeologists I’ve spoken with find the Eastern Levant hypothesis origin for the Mushabian culture more convincing. Interestingly, though, the Mushabian (and Ramonian) appear to have actually been present in the Nile Delta during the Epipaleolithic, but the explanation I’ve seen as to what this represents is that it’s a very ancient extension of the Mushabian and Ramonian territory rather than a North African holdover before some supposed expansion into the Levant.

Anyway, I ask if anyone can hazard a guess, because, for example, Natufian hunter-gatherers had a significant minority of Iberomaurusian ancestry (and apparently some AEA ancestry), seem to have shown at least one Iberomaurusian-inspired practice in the form of tooth avulsion, and, again, carried E-Z830 lineages (maybe an unrecognized minority of E-M78 lineages, too?), all of which show some kind of relationship to Africa. In spite of that, finding a cultural correlate that might have brought these all to the Levant hasn’t been simple for me to identify.

The potential signals of contact between North Africa and the Levant during the Epipaleoithic seem to only be in the form of art and slightly vague similarities in assemblages from the early Epipaleolithic, though not during the middle and late Epipaleolithic, which would be most relevant in this case. I am aware that there can technically be population introgression without many signs of cultural or archaeological changes, which one of the publications I’ve looked at briefly entertains. Although, I tend to see that as a last resort explanation that’s worth entertaining if the data helps point one toward that conclusion.