Basal Eurasian discussion

[Jerome]

@Kale @kolompar

What do you think of India as a hub for early Eurasian population? I have seen that many haplogroup migration maps locate the origin of F in India, and India seems to have some unresolved F*. If this would be true, the main OoA bottleneck could be in there. IJ and G (west/basal) headed west first, then C, D1, F2, H and K stayed here and shared extra bottleneck for about 3,000 years. India is on a similar temperature zone of the Arabian peninsula, and the OoA allele selection may be possible.

However, if India is not the hub, then there may be some issues on the "Arabian Standstill". Do you agree that this selection/OoA bottleneck had already happened within Africa, and there was no real standstill after the exodus out of Africa (55-50kya)? If so, maybe all OoA related elements (pre-C, pre-F, pre-D1, etc.) were taken over by the E-population later.

Not sure how exactly this falls in to the debate, but I've always found it interesting that there is minimal Haplogroup DE among South Asian populations (despite being adjacent to geographical regions carrying Haplogroups D and E).

I think aasi was C and F(if you look at tribal distribution in south asia).
H was probably not original to the region but arrived there early(a bit before the LGM).

The C and F is similar to UST ishim/crown Eurasian type people and was the likely the ydna of non-basal eurasians.

[Desdonas]

@Jerome @Merriku

A phylogeny tree of H (time based on ftdna or Yfull)

H (L901) (45k/45.6k)
H2-P96 (Near East)
H1'3-M2826 (45k/43.1k)
H1-L902 (41k)
H1a-M69 (39k/38.7k, South Asia)
H1b-B108 (a Myanmar sample)
H3-Z5857 (South Asia, 30k/28k)

Also, UP West Eurasia has no H-M2826 specimen. So it's hard to imagine that H-M2826 is not AASI or arrived at South Asia after LGM. The B108 sample may represent some prehistoric or recent contact between South Asia and Southeast Asia.

Since M2826 is most likely AASI-specific, its direct ancestor H (L901) may represent an ancient Eurasian lineage in 46-45kya which contributes to both AASI and Ancient Near Eastern populations. The origin of H is most likely between West Asia and core-India (thus Makran, Indus, etc.).

Last, I once thought that there are some unresolved F* in India. But now I change my view and consider that they are based on old researches. During that period, H3 was not popular and the researches may neglect that.

[Merriku]

@Jerome @Merriku

A phylogeny tree of H (time based on ftdna or Yfull)

H (L901) (45k/45.6k)
  H2-P96 (Near East)
  H1'3-M2826 (45k/43.1k)
      H1-L902 (41k)
          H1a-M69 (39k/38.7k, South Asia)
          H1b-B108 (a Myanmar sample)
      H3-Z5857 (South Asia, 30k/28k)

Also, UP West Eurasia has no H-M2826 specimen. So it's hard to imagine that H-M2826 is not AASI or arrived at South Asia after LGM. The B108 sample may represent some prehistoric or recent contact between South Asia and Southeast Asia.

Since M2826 is most likely AASI-specific, its direct ancestor H (L901) may represent an ancient Eurasian lineage in 46-45kya which contributes to both AASI and Ancient Near Eastern populations. The origin of H is most likely between West Asia and core-India (thus Makran, Indus, etc.).

Last, I once thought that there are some unresolved F* in India. But now I change my view and consider that they are based on old researches. During that period, H3 was not popular and the researches may neglect that.

That Burmese sample (burm10) is intriguing for sure. If it is truly basal H1 then it would suggest that H1 was lost in SE Asia (and maybe even east of the South Asia as a whole) to founder effects and that it is one the lucky relict subclades. 

I ran its vcf on snipsa and it can be resolved downstream of H1-
H-Z13966 YTree (yfull.com)
But this subclade prediction is based on a single positive SNP downstream of H-M69 level.

Analysis with the BAM file can provide much more reliable  subclade prediction. But unfortunately raw data from this study is available in some unusual format (the format used by Complete Genomics).

[Norfern-Ostrobothnian]

I highly doubt that there are truly ancestral forms of any haplogroup in the present day, because that would require absolutely no mutations to take place in the haploid region which makes absolutely no sense. It is rather an earlier split than the others is what I'd think it is, where it's novel mutations simply haven't been found elsewhere and that's why it appears as if is actually ancestral. I would also not put much weight onto these stray early splits in South and Southeast Asia since the population size elevates the odds of some minor lineages persisting there longer than elsewhere and because the haplotype itself may have wandered about from somewhere else, not to mention that the predominant ancestry of the Burmese doesn't come from the Southeast Asian region in Upper Paleolithic terms.

[Southpaw]

There was a time when Mechta-Afalou and Taforalt-like individuals from Northern Africa were supposed to be Cromagnoids from Europe who crossed the Gibraltar.

Anthropologist thought there were resemblance with Upper Paleolithic skulls from Europe due to robust skeletons. I guess it was just convergent evolution.

Now, i wonder whether ANA is a real autosomal and Basal Eurasian is real as well or the story behind both of them is more complex or both Basal Eurasian and ANA ultimately stems from a common shared ancestor which was admixed with different set of people.

In my personal opinion, i don't really think Y-DNA E came from Asia, i think Ethiopian Highlands is the best bet, E-M215 pushed to North Africa and colonized the whole region with some spinoffs migrating to Levant and mixing with local populations to create the Natufians we know, i might make a guesstimation that Northern Africa prior to E-M215 might have been populated by people resembling somewhat more the Eurasians, hence the mtDNA U6a among Iberomaurusians.

[Merriku]

I highly doubt that there are truly ancestral forms of any haplogroup in the present day, because that would require absolutely no mutations to take place in the haploid region which makes absolutely no sense. It is rather an earlier split than the others is what I'd think it is, where it's novel mutations simply haven't been found elsewhere and that's why it appears as if is actually ancestral. I would also not put much weight onto these stray early splits in South and Southeast Asia since the population size elevates the odds of some minor lineages persisting there longer than elsewhere and because the haplotype itself may have wandered about from somewhere else, not to mention that the predominant ancestry of the Burmese doesn't come from the Southeast Asian region in Upper Paleolithic terms.

That's true, if it truly is unresolved at H1 level then it will have 40k years worth of novel SNPs

[Hammas]

They can be modeled in G25 with 26.5% Paleolithic European ancestry though, but I don’t know how accurate this model is

MAR Taforalt ► Average   
Fit 30.84
39% Dinka
33% Pinarbasi HG
26.5% Pestera Muierii EUP
1.5% Yoruba

[Southpaw]

The way the model can be interpreted is that ANA in terms of deep split stands somewhere between SSA and Eurasian, according to the tree provided by Lazaridis it split from a common ancestor of ANA - Core Eurasian - Basal Eurasian, with ANA splitting earlier and then Core Eurasian from Basal Eurasians.

Hence these models without context are useless.

[Desdonas]

I highly doubt that there are truly ancestral forms of any haplogroup in the present day, because that would require absolutely no mutations to take place in the haploid region which makes absolutely no sense. It is rather an earlier split than the others is what I'd think it is, where it's novel mutations simply haven't been found elsewhere and that's why it appears as if is actually ancestral. I would also not put much weight onto these stray early splits in South and Southeast Asia since the population size elevates the odds of some minor lineages persisting there longer than elsewhere and because the haplotype itself may have wandered about from somewhere else, not to mention that the predominant ancestry of the Burmese doesn't come from the Southeast Asian region in Upper Paleolithic terms.

That's true, if it truly is unresolved at H1 level then it will have 40k years worth of novel SNPs

Although there may be a 41k split of Myanmar H1 and South Asian H-M69, the other two splits between AASI and its SEA counterparts are 45-44k. The H1-L902 level still has a bottleneck of 34 snps.

https://discover.familytreedna.com/y-dna...scientific

https://discover.familytreedna.com/y-dna/K-Y28394/story
https://discover.familytreedna.com/y-dna/C-B66/story

3,000 years may be significant for the East Eurasian meta-population, since their expansion was very rapidly. So when the basal B108 sample split from M69, Australasians may have crossed Wallacea, and Hoabinhian and the major ancestor of East Asians were already in MSEA or Yunnan. In this case, the B108 sample still represents a secondary movement into SEA.

[Jerome]

@Jerome @Merriku

A phylogeny tree of H (time based on ftdna or Yfull)

H (L901) (45k/45.6k)
  H2-P96 (Near East)
  H1'3-M2826 (45k/43.1k)
      H1-L902 (41k)
          H1a-M69 (39k/38.7k, South Asia)
          H1b-B108 (a Myanmar sample)
      H3-Z5857 (South Asia, 30k/28k)

Also, UP West Eurasia has no H-M2826 specimen. So it's hard to imagine that H-M2826 is not AASI or arrived at South Asia after LGM. The B108 sample may represent some prehistoric or recent contact between South Asia and Southeast Asia.

Since M2826 is most likely AASI-specific, its direct ancestor H (L901) may represent an ancient Eurasian lineage in 46-45kya which contributes to both AASI and Ancient Near Eastern populations. The origin of H is most likely between West Asia and core-India (thus Makran, Indus, etc.).

Last, I once thought that there are some unresolved F* in India. But now I change my view and consider that they are based on old researches. During that period, H3 was not popular and the researches may neglect that.

I already mentioned that It reaches south asia before LGM.

There's a catalhoyuk sample with H3,I don't know if it's high coverage or not.

[Jerome]

There was a time when Mechta-Afalou and Taforalt-like individuals from Northern Africa were supposed to be Cromagnoids from Europe who crossed the Gibraltar.

Anthropologist thought there were resemblance with Upper Paleolithic skulls from Europe due to robust skeletons. I guess it was just convergent evolution.

Now, i wonder whether ANA is a real autosomal and Basal Eurasian is real as well or the story behind both of them is more complex or both Basal Eurasian and ANA ultimately stems from a common shared ancestor which was admixed with different set of people.

In my personal opinion, i don't really think Y-DNA E came from Asia, i think Ethiopian Highlands is the best bet, E-M215 pushed to North Africa and colonized the whole region with some spinoffs migrating to Levant and mixing with local populations to create the Natufians we know, i might make a guesstimation that Northern Africa prior to E-M215 might have been populated by people resembling somewhat more the Eurasians, hence the mtDNA U6a among Iberomaurusians.

If I had to say it would be Nile valley...


ANA is kind of a proto-eurasian(it splits from the node that gives rise to eurasians rather than splitting from an earlier African node),so that really complicates the non Eurasian origin of E.

I don't think it came from Asia,but rather something like the boundary,like the nile valley,Sinai.

It could have born in north of the Ethiopian highlands,but in an Eurasian like population that went extinct later?

It's difficult to say,but the fact that core African populations which split earlier don't really have E and in which populations that do carrry,the TMRCA and formation date is very late,(For example E starts splitting around 53k BC,that's more in line with the timeline of Basal Eurasian like populations).

[Kale]

There's a catalhoyuk sample with H3,I don't know if it's high coverage or not.

All of the Catalhoyuk are poor coverage. I can't give you a source, but I'm 99% sure someone did the digging and it was not actually H3, but something mundane (maybe H2?).

[Jerome]

Did somebody figure out what dzuduana is?
It seems radically different than what Lazaridis projected it as(similar to anatolia_hg).

From rough modelling,it seems to be a mix of aurignacian+Proto-natufian/Kebaran+(the non-ANE part of iran_n)+Proto-WHG.

Is this right or is it something even different?

And from modelling CHG it seems there's barely any continuity between CHG and dzuduana,and their lithuc industry difference denotes this.

CHG comes out as 10-20% Dzuduana+5% EHG+Rest Zarzian/Iran_n.

Satsurblia seems to be the least dzuduana(5%) while kotias has more(20-25%).

If we use something like Hotu as a source population which is more ANE rich then the need for EHG almost disappears.

[Kale]

My current hypothesis regarding Dzudzuana
1) Goyet/Fournol represent unadmixed 'West Eurasians'
2) Post-CI Eastern Europe get a small pulse of Basal from the Caucasus. Descendents of this mix include Kostenki/Sunghir, the Western part of ANE, and 'CWE'.
3) Dzudzuana = The above basal + 'CWE'

[TanTin]

Did somebody figure out what dzuduana is?
It seems radically different than what Lazaridis projected it as(similar to anatolia_hg).

From rough modelling,it seems to be a mix of aurignacian+Proto-natufian/Kebaran+(the non-ANE part of iran_n)+Proto-WHG.

Is this right or is it something even different?

And from modelling CHG it seems there's barely any continuity between CHG and dzuduana,and their lithuc industry difference denotes this.

CHG comes out as 10-20% Dzuduana+5% EHG+Rest Zarzian/Iran_n.

Satsurblia seems to be the least dzuduana(5%) while kotias has more(20-25%).

If we use something like Hotu as a source population which is more ANE rich then the need for EHG almost disappears.

These days I am still too far from Dz and IE.
I try to figure out how the first HS appeared from Deni and Neand.
It is weird the whole variaty that we see in Africa, then it goes through some bottleneck and afther that few main lines continue.
But the most misterious are some archaic groups or individuals that show close relation to something like Marmoset and Macaque.. And these archaic rich individuals do not point to Africa, but to somewhere else.  Even America.
I am going to include more individuals for my next tests, including Dz. 
So far I have some idea, but still lot of mess. Fortunately there are some clear parts of the picture. There are some clear migration ways, where we may follow the connection between Africa and OOA. 
Another big challange are the haplogroups:  we see the oldest haplogropus in Africa, but I find some oldest archaic markers outside of Africa. 
I was pretty much able to find the missing components for WHG / EHG..  However these are archaic.. That means they have lot of markers from groups like: Denisova, Gorilla, Neand.. etc.
I am still focused on Africa,  I want to make sure what exactly happened there, how these archaic hominoids mixed between each other so as a result we have HS ? Why do we see more Deni and Neand outside of Africa ? Was this direct admix, or random distribution during OOA ?  All kind of questions and each of these is a new door to open..