HV1b2 Origin Hypotheses

[SUPREEEEEME]

From what we know, HV1b2 is almost exclusively found in Ashkenazi Jews at a frequency of ~3%-3.54%, and Europeans of maternal Jewish descent. I am of the opinion that there are 3 potential origins for HV1b2, and I will explain the "evidence" in each case:

Theory 1: Levantine/Canaanite
A Levantine/Canaanite origin is supported by the presence of HV1b2 in a Romaniote Jew (from Larisa) - indicating the possibility of coming from an ancestral source. This is furthered by a study that showed that HV1b2 is very closely related to a lineage in Iraqi and Georgian Jews (the study seems to have been removed - I can no longer find it).

Additionally, HV1a'b'c (upstream of HV1b2) is present in a sample from the Late Chalcolithic (3800 BCE-4500 BCE) Israel. Similarly, HV1b1 is present in a Canaanite sample from Lebanon, 1650 BCE-1750 BCE. The reason that I'm bringing these samples up is, in my mind, while not HV1b2, since these subclades were found in ancient Levantine remains, there's no reason why HV1b2 was not present there as well (which isn't the strongest of arguments, admittedly).

Theory 2: Ancient Egyptian
The only evidence for this origin are the only ancient sample(s) of HV1b2 - which are from Egypt. From the actual study, there's only one sample, dated to the Ptolemaic period 545 BCE - 756 BCE. According to this blog (http://www.razib.com/wordpress/category/egypt/), there were two HV1b2 mummies.

Theory 3: Northern Mesopotamian
This is supported by HV1b2's position between Armenian and Assyrian branches. This origin is also supported by a study from last year (2019), which I'll link. Additionally, HV1b2 has been found in a Yezidi Kurd and an Armenian, which could support it.

https://media.springernature.com/full/springer-static/image/art%3A10.1038%2Fs41598-019-48596-1/MediaObjects/41598_2019_48596_Fig2_HTML.png?as=webp
https://media.springernature.com/lw685/springer-static/image/art%3A10.1038%2Fncomms3543/MediaObjects/41467_2013_Article_BFncomms3543_Fig5_HTML.jpg?as=w ebp

Hopefully in the not too distant future, there will be more cases (ancient and modern) of HV1b2.

Sources:
https://www.nature.com/articles/ncomms3543
https://academic.oup.com/gbe/article/8/4/1132/2574015 - although, it's by Elhaik - so not a trusted source
https://www.yfull.com/mtree/HV1b2/
https://www.nature.com/articles/s41467-018-05649-9#Sec22
https://www.biorxiv.org/content/10.1101/142448v1.supplementary-material
https://www.nature.com/articles/ncomms15694#Sec30
https://lists.rootsweb.com/hyperkitty/list/genealogy-dna.rootsweb.com/thread/35479413/
https://www.nature.com/articles/s41598-019-48596-1
http://kurdishdna.blogspot.com/2013/10/hv1b2-discovered-among-yezidi-kurd.html

[StillWater]

Good post.

A few points:

- some sources indicate that the frequency of HV1b2 among Ashkenazim is lower and closer to 1%
- the presence in Egyptian mummies adds more evidence to a Levantine origin, as opposed to evoking a new source. It may have arrived in Israel from Egypt, but given the clade's MRCA date of 2000ybp (we can double check this), it would've entered Jews through an Israelite/Judean intermediary. However, if forced to argue a strictly Egyptian origin, one can point out that the parent clade HV1b-T152C! (this is deeper than HV1b) does exist in a Tunisian (uncertain of his ethnicity, if anyone can find the exact record - I will appreciate it). Likewise, HV1b-T152C! exists in Assyrians, adding support for a Mesopotamian origin, and it exists in a Yemeni, now adding support for a Levantine origin. Given how weird FTDNA's haplotree is, the Yemeni could be further downstream and maybe even at HV1b2 - would be helpful if someone could find his record somewhere. FTDNA's haplotree does list someone from Azerbaijan under HV1b2, but this one could even be Ashkenazi(plenty of Ashkenazim lived there). The other subclades of HV1b-T152C! belong to Assyrian,Armenians, Italians, and Saudis, and Iranians.
- it should be noted (and it was originally pointed out to me by Agamemnon) that this clade is disproportionally found among Litvaks. From what I've seen, its intra-Ashkenazi distribution isn't only disproportional, but overwhelmingly Litvish. However, FTDNA's halpotree does list a lot from Poland (based on what I've seen from detailed records, these are likely to be from the Litvak areas of Poland and/or Western Belarus).
- The Armenian you're referring to isn't a "Derenko Armenian". Derenko is the study's author. The Armenian HV1b2 was found in Mush, Eastern Turkey, in an area with almost no Armenians left. All the Armenians there are Crypto-Armenians, I think. He's the one with the Turkish flag on YFull.
- In Costa's big MTDNA study, they didn't find any Western Ashkenazim with HV1b2, though haplotree does list some. As I've said before, I suspect this to be a Knaani lineage: coming through Byzantium/Greece/Black Sea Greek colonies and moving into Knaan Yevan (Northeastern Europe). I use the Romaniote with HV1b2 as support for this theory.
- Elhaik is an imbecile. He examined HV1b2 as an MTDNA found among Cohenim. Cohen status is patrilineal, but he used the diversity of MTDNA among Cohenim to imply the nonexistence of a Cohen lineage.
- While it has been found in a Yezidi Kurd (one who manages to get himself banned from every forum, so I can't ask him for more details), the Kurdish blog post is Elhaik-nonsense. I'm hoping there are enough mutations to compare his result to that of Jews.
- The study which argues for a Mesopotamian origin, also implies that HV1b2 has been found in the Levant, but their source for it is unclear. If someone can clarify this, it would be really helpful.
- This may prove very helpful :http://генофонд.рф/data/mtdna/the_tree/hg/ages.htm# (I'm just too tired to read it). If can provide a legend/guide how to interpret each column, it would help.

HV1b2 HV1b 4 0.2500 +/- 0.2500 835.7500 +/- 835.7500 0.2500 +/- 0.2500 1291.7500 +/- 1291.7500 0.2500 +/- 0.2500 1997.5000 +/- 1997.5000

Here is the data to compare the Kurdish and Armenian HV1b2 to Ashkenazim:

http://www.ianlogan.co.uk/sequences_by_group/hv1_genbank_sequences.htm

Can anyone explain how different/similar they are? The Armenian seems to be missing the T52C mutation, while the Kurd either tested at a very low resolution or is radically different from all others.

[StillWater]

Big Update: Have asked 2 Mountain Jews to look through their matches. HV1b2 seems to be reasonably common Mountain Jews. One had roughly 0.6% Mountain Jewish HV1b2s (8 full Mountain Jews and one that was 7/8 Mountain Jewish and 1/8 gentile Eastern European) and 0.55% Ashkenazi HV1b2s. I haven't yet received statistics from the other. When the maternal grandmother was localized, Quba (Azerbaijan) seemed to constantly reappear. Given its recurrent presence in Quba, the frequency of HV1b2 might've been greater had I asked Mountain Jews from Azerbaijan. I've seen that my local Ashkenazi origins affect the frequency I see of various lineages and admixtures relative to others' matches.

This greatly reduces the Egyptian hypothesis as an independent one and shifts most of the remaining weight to Levantine vs Mespotamian. Given the presence of HV1b2 among Mountain Jews, I have more confidence that Costa et al 2013 were in fact precisely speaking about HV1b2 when stating that a related lineage existed in Georgian and Iraqi Jews. Currently, the two groups with highest known frequency of HV1b2 are Ashkenazim (particularly, Litvaks) and Mountain Jews.

[Nebro]

HV1b1 was found on kenya Pastoral Neolithic , if that could hint on something ...

[SUPREEEEEME]

Big Update: Have asked 2 Mountain Jews to look through their matches. HV1b2 seems to be reasonably common Mountain Jews. One had roughly 0.6% Mountain Jewish HV1b2s (8 full Mountain Jews and one that was 7/8 Mountain Jewish and 1/8 gentile Eastern European) and 0.55% Ashkenazi HV1b2s. I haven't yet received statistics from the other. When the maternal grandmother was localized, Quba (Azerbaijan) seemed to constantly reappear. Given its recurrent presence in Quba, the frequency of HV1b2 might've been greater had I asked Mountain Jews from Azerbaijan. I've seen that my local Ashkenazi origins affect the frequency I see of various lineages and admixtures relative to others' matches.

This greatly reduces the Egyptian hypothesis as an independent one and shifts most of the remaining weight to Levantine vs Mespotamian. Given the presence of HV1b2 among Mountain Jews, I have more confidence that Costa et al 2013 were in fact precisely speaking about HV1b2 when stating that a related lineage existed in Georgian and Iraqi Jews. Currently, the two groups with highest known frequency of HV1b2 are Ashkenazim (particularly, Litvaks) and Mountain Jews.

Thanks for the info!

[Nebro]

imo if HV1b1 was in ancient kushitis that means HV1b2 is either egyptian or levantinese but not any far eastern...

[SUPREEEEEME]

We have some relevant ancient samples!

From Skourtanioti et al:
ALA018: HV1b3b - Alalakh

From Agranat-Tamir et al:
I3965: HV1b3 - Hazor

Both HV1b3 and HV1b2 fall under HV1b-a:
https://www.yfull.com/mtree/HV1b-a/

This is how it looks like in terms of ancient samples:
- HV1b-a
-- HV1b2 --> Ptolemaic Era Egyptian
-- HV1b3 --> Hazor Canaanite
--- HV1b3b --> Alalakh Amorite

In my opinion, this supports a Levantine origin for HV1b2, if not all of HV1b-a.