New blogs

Leherensuge was replaced in October 2010 by two new blogs: For what they were... we are and For what we are... they will be. Check them out.

Monday, June 9, 2008

The Y-DNA spread map (Eurasia)


As the previous map sparked some discussion on Y-DNA spread, here there is a map I already had about it:




Macro-haplogroups are color coded:
- Red: F, magenta: K
- Yellow: C
- Cyan blue: D
- White: root clades CT and CF

The electric blue thick line approaches the max. extent of ice cover (in Eurasia only)

5 comments:

terryt said...

Maju. I remember reading a post a couple of hours ago regarding dating of Y-hap O. I haven't got time to search it out now but this could be a good place to put the following: an approximate timetable for the Y-hap evolution, based on a sort of round-figured acceptance of all the evidence at face value.

The A-T clade had diversified by 80,000 years ago and one branch became B. The clade C-T branched off this by 70,000. This C-T clade then split into two (D/E and C/F-T) by 60,000 years ago. The C and F clades both began diversifying about 50,000 years ago. K diversified from about 40,000. NO and QR began diversifying about 30,000 years ago. Also IJ.

I have certainly come to accept that your knowledge of the trees from here on is well researched and very likely to be correct.

Maju said...

The O3-subclades article is Hong Shi et al, 'Y-Chromosome Evidence of Southern Origin of the East Asian–Specific Haplogroup O3-M122'. AJHG, 2005 (http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1226206). There was also a more recent paper, mentioned at Dienekes that gave as well high age estimates to O2 subclades in Hainan aborigines.

Hong Shi median estimates are between 25 and 30,000 BP, depending on the specific subclade, but the error margins range between 17 and 39,000 BP.


The A-T clade had diversified by 80,000 years ago

I don't think you can give such recent estimates for Y-DNA Adam, really. As said before the normal estimates are over 90 KY and, IMO, it could be even double that.

... and one branch became B. The clade C-T branched off this by 70,000. This C-T clade then split into two (D/E and C/F-T) by 60,000 years ago. The C and F clades both began diversifying about 50,000 years ago. K diversified from about 40,000. NO and QR began diversifying about 30,000 years ago. Also IJ.

If you want to see it that way...

IMO, the SNP distances in the best studied branches (R1, O3, IJ, E) suggest much more deeper ages for each single node. The arguable (and yet to be refined) archaeological continuity in South Asia might well also weight in favor of an older Sapiens settlement in that region.

It's not clear but I will eventually write something on that because I make little sense of such recent dates for OOA and out-of-South-Asia clades when their third ot fourth level derivates are so old.

Normally the argumentation (as in Karafet 2008) is the opposite: IF OOA was c. 70,000 BP, THEN clades like R1 or O3 are younger than 20,000 BP, and their subclades would be almost Neolithic. I am applying now the inverse logic: IF O3a is 30,000 years old, THEN OOA (and the CT>CF transition) are probably older than 100,000 BP (you should add like 100% or more to all dates).

I can't say much more, make your own calculations and see if what I think makes sense to you too. The ISOGG tree has already been actualized with all the new mutations of Karafet-2008 (and others), so the most recent knowledge in Y-DNA lineages and its SNPs is there for anyone to use it freely.

There are three problems to use SNP counting to estimate dates:

1. Different branches have very different levels of knowledge, and this is clearly expressed in the number of known mutations. In principle all branches should be about the same length in SNPs but right now some "have" more than 60 SNPs while others just a handful. The longest branch is in R1b1 subclades, what is obviously caused by some Eurocentrism in genetic studies. I am using the longest available branch to make my estimates, as there are obviously many "mising" SNPs in the others (and therefore should be apportioned).

2. Mutation rates. In principle the bigger the population the fastest mutations happen but the less likely they become fixated (they form distinct haplogroups). Small populations should be genetically "conservative" (lower "gross" mutation rate) but then also new mutations are also more likely to become fixated within small populations as well and often should mean founder effects. By the moment I've "solved" this inconvenient uncertainty by using constant mutation rates independently of any population estimates (necesarily a hunch anyhow). Not the best solution, I am aware.

3. We really know little or nothing this way on the SNP diversity that may have arisen since the latest demographic expansions (Neolithic, Epipaleolithic, post-LGM) because these are not likely to constitute easily recognizable haplogroups, or sufficiently large groups to be acknowledged as haplogroups. In this aspect we are dependent on TRMCA or archaeologically absed estimates, which may be correct or wrong.

There's a lot of uncertainty as a whole trying to estimate haplogroup ages. We do know the temporal direction of divergence and have some clues on the genetic distance between nodes but we realy do not have easy answers for their ages.

The timing of this expansion and diversification therefore cannot be independent of Archaeology but in many places Archaeology also gives limited clues.

And that's how it is.

I have certainly come to accept that your knowledge of the trees from here on is well researched and very likely to be correct.

I do have some good idea, I understand, but I can certainly comit errors, specially in the mtDNA area, where knowledge is spread around in many papers and there is no single "official" centralized site to recopilate and standarize all that, like happens with Y-DNA (http://www.isogg.org/tree/index.html).

terryt said...

I agree the dates are extremely ill defined. And I made the very unlikely assumption that each diversification took the same length of time.

I remember you posted a link to a very good mtDNA diagram somewhere. I'm presuming for now it's the link you've provided here.

terryt said...

Of course it's not is it. You've posted a link to Y-chromosome. Do you remember thet link to the mtDNA tree?

Maju said...

Well for the mtDNA maps below I worked with the many trees (a little complicated, I know) in http://www.ianlogan.co.uk/mtdna.htm. But this is not waht you're asking for.

You must mean http://www.mitomap.org/mitomap-phylogeny.pdf but notice it's incomplete. Check the above link for more details, specially for the M macro-haplogroup, ignored for so long.