March 01, 2013

Extremely old (237–581 kya) root of human Y-chromosome phylogeny

I had mentioned this research before, and now it has officially been published. There are three things to be excited about this new paper:

First, it forces us to consider the possibility of either (i) archaic admixture in Africa, or (ii) a much more ancient time depth of modern humans than the first fossils from Ethiopia dated to about ~200 thousand years ago.

Second, it underscores the importance of collaboration between academia and regular folk, since it was the combined contributions of academics, genetic genealogists, and the owners of the new A00 basal Y-chromosomes that made this new discovery possible.

And, third, it shows that the extraordinary can be discovered without contacting isolated tribes or seeking human bones in remote regions, but rather through careful scrutiny of large volumes of data for the proverbial needle in the haystack.

The paper developed a model of Y-chromosome mutation based on the estimate of Kong et al. Significantly, though:

If we were to use the higher mutation rate (1.0x10^-9 per base per year6) rather than a realistic range derived from whole-genome sequencing (4.39x10^-9  -   7.07x10^- 9), the estimated TMRCA for the tree incorporating A00 as the basal lineage would be 209 kya, which is only slightly older than current estimates of the TMRCA of mtDNA and the age of the oldest AMH fossil remains. We note, however, that the higher mutation rate produces an estimate for the common ancestor of all non-African Y chromosome haplogroups (C through T) of ~39 kya6 (i.e., versus ~63 kya for the mutation rate used here).
A 39kya common ancestor for Eurasian Y-chromosomes makes no sense, since we now know for sure that by that time, the differentiation of Eurasians was already well on its way and modern humans in remote parts of the Old World have been documented much earlier than that time.

A ~63kya common ancestor, on the other hand, fits nicely with my "two deserts" theory of modern human origins, according to which the ancestors of Eurasians faced an ecological crisis in Arabia when it became much drier post-70kya; that seems like a most opportune time for the major Eurasian bottleneck and the corresponding coalescence of Eurasian Y-chromosomes to a single man. And, while there is no a priori reason for Y chromosomes and mtDNA to behave similarly, the age of the "older" Eurasian ur-mother, haplogroup N at 59 thousand years, with presumably an older ancestor within mtDNA haplogroup L3 founding the Eurasian population.

Also, if modern human-Neandertal admixture had occurred  "most likely 47,000–65,000 years ago", then the expansion of modern Proto-Eurasians within a 70-60kya timeframe north Out-of-Arabia would have brought them in contact with their northern Neandertal neighbors. On the other hand, it would be incredible if modern humans experienced admixture with Neandertals but were still much later a very small population (to allow for the coalescence of their Y-chromosomes to one man ~39kya).


So, in summary, the mutation rate used by the authors seems consistent with what we know about an important calibration point of the human story.

But, who were the people in Africa responsible for the introgression of A00 chromosomes? Mendez et al. used the haplotype of the African American A00 individual and discovered his patrilineal kin among the Mbo of Cameroon, who are Bantu farmers. 

I have observed before that Pygmies and Bushmen represent only a tiny fraction of pre-existing African genetic diversity, the part that had not yet been absorbed into the farmers' expanding population by the time that Africa came to the attention of of modern science. We see traces of Pygmy and Bushman ancestry in some African farmers, and there were probably other groups, no longer extant as distinct ethno-cultural entities, but, nonetheless, surviving as genetic fragments in the genomes of the farmers.

Thus, while it still makes sense to study the surviving hunter-gatherers of Africa who make up perhaps a percent or less of the population of Africa, it may be equally important to study different groups of African farmers who may possess a much richer treasure trove (albeit diluted) of such "Palaeoafrican" ancestry.

Finally:

Although the stochastic nature of the evolutionary process can explain the aforementioned incongruences, the extreme age and rarity of the A00 lineage point to the possibility of a highly structured ancestral population, consistent with recent work on the autosomes.40,41,43,44 This could take the form of long-standing population structure among AMH populations45 or archaic introgression from an archaic form into the ancestors of AMHs.46 Interestingly, the Mbo live less than 800 km away from a Nigerian site known as Iwo Eleru, where human skeletal remains with both archaic and modern features were found and dated to ~13 kya.47 Further surveys in sub- Saharan Africa and in the African Diaspora might uncover more diverged basal lineages, which will help to disentangle some of the complex evolutionary processes that shape patterns of Y chromosome diversity.


AJHG 10.1016/j.ajhg.2013.02.002

An African American Paternal Lineage Adds an Extremely Ancient Root to the Human Y Chromosome Phylogenetic Tree 

Fernando L. Mendez et al.

We report the discovery of an African American Y chromosome that carries the ancestral state of all SNPs that defined the basal portion of the Y chromosome phylogenetic tree. We sequenced ∼240 kb of this chromosome to identify private, derived mutations on this lineage, which we named A00. We then estimated the time to the most recent common ancestor (TMRCA) for the Y tree as 338 thousand years ago (kya) (95% confidence interval = 237–581 kya). Remarkably, this exceeds current estimates of the mtDNA TMRCA, as well as those of the age of the oldest anatomically modern human fossils. The extremely ancient age combined with the rarity of the A00 lineage, which we also find at very low frequency in central Africa, point to the importance of considering more complex models for the origin of Y chromosome diversity. These models include ancient population structure and the possibility of archaic introgression of Y chromosomes into anatomically modern humans. The A00 lineage was discovered in a large database of consumer samples of African Americans and has not been identified in traditional hunter-gatherer populations from sub-Saharan Africa. This underscores how the stochastic nature of the genealogical process can affect inference from a single locus and warrants caution during the interpretation of the geographic location of divergent branches of the Y chromosome phylogenetic tree for the elucidation of human origins.

Link

23 comments:

Mike Keesey said...

It seems also quite possible that anatomically modern humans are just a lot older than we think, thanks to the Signor-Lipps effect. (And consider how terrible the fossil record in Central Africa is.)

Although given what we now know about "archaic" admixture, introgression of A00 from an "archaic" populations also seems quite possible.

How can we test these two alternatives? Identifying the genes that cause our high foreheads, etc.?

BES said...

Thanks for your comments, Dienekes. I too picked up on the implications for the timing out-of-Africa migration as the most convincing evidence for this new mutation rate (see my FB posting tonight, before reading your blog).

Another thing I might mention is that our African American project members are a far closer match to the Bangwa people, whose haplotypes are found in the SMGF database -- if you can ever get in there and search it now, with all the restrictions being placed on it -- than to the Mbo who were sequenced. The Bangwa and Mbo are close neighbors, and rivals.

It's good to have as much genetic distance as possible between the A00 studied, to get some idea of their variance. But when talking about who our project member's kinfolk in Africa might be, it's a far better chance they're Bangwa, from Fontem, Lebialem, than Mbo. We found two SNPs in our project members that were negative in the Mbo -- perhaps they'd be found positive if Bangwa were tested.

It's also likely that a few other peoples in that region may harbor more A00. It would be useful if UCL would reveal which ethnic groups they've already sampled, where there were no matches, so we'd know where not to look.

This paper had a narrow focus, but we have much richer data, with multiple 'new' ancient branches found (even if they aren't as ancient as A00), that haven't yet been fully described. There is still more to be learned from these about African population history.

Bonnie Schrack

shenandoah said...

"...it forces us to consider the possibility of either (i) archaic admixture in Africa, or (ii) a much more ancient time depth of modern humans than the first fossils from Ethiopia dated to about ~200 thousand years ago..."

Or, (iii) both factors.

Also, as a humble layperson, it is difficult for me to understand how the ~antiquity of haplotypes or groups is calculated: what do they compare them with?

Do they compare all Human yDNA testing results with some particular standard model genome, perhaps that of our 'common ancestor', the Chimpanzee?

Or are they calculated strictly by comparison with other known Human yDNA? I'm sorry, but could someone please answer that question for me, if anyone knows? I've been puzzled about that point forever, lol.

eurologist said...

Mike,

Another data point is from Neanderthal/ modern human divergence estimates. I still don't trust datings too much, but if this is of the order of 400,000 years, it would be hard to believe that fully AMHs began to appear just 60,000 years later.

Conversely, the new date seems to be roughly from the period when Neanderthal/heidelbergensis last mixed with with Africans (before ooA). I'd love to see how divergent the y-DNA of those were.

Orang More said...

If it is due to admixture with archaic hominids, it is interesting that, unlike for neanderthals and denisovans, the Y-chromosome remained in the population at significant frequencies, does this mean that the level of admixture (autosomal) was much higher? Could the high genetic diversity within africa be due to high admixture with preexisting hominids?

Mike Keesey said...

@shenandoah, chimpanzees are our extant sister group, not our 'common ancestor'.

@eurologist, good point (with a well-stated caveat).

BES said...

Hi Shenandoah,

It's pretty technical and a subject of great controversy -- I hope others will comment on dating, but first of all it's done by using an estimated mutation rate for how often SNPs occur, and then counting the SNPs from a known point back through the phylogeny to the earliest branching point.

The Chimp and Gorilla sequences were used largely to determine which was the ancestral state of each SNP, and which was derived.

Since they had lots of STR data -- thanks to the excellent work of Thomas Krahn and his colleagues -- they also used STR dating, but just for the question of how long it had been since the African American lineage separated from the Mbo lineages.

Now, as to Orang's comments, I would not say that we could consider this Y chromosome lineage to be remaining in the population at significant frequencies. It's extremely rare. There is no evidence at all of high levels of autosomal admixture. What is reasonable to consider as a hypothesis is that small numbers of archaic humans probably survived longer in Africa than in Eurasia, given that it was their original homeland, containing plenty of food resources compared with Ice Age conditions in the North, and dense forest which could serve as a refuge.

It's necessary in a paper like this to recognize all the plausible alternatives, but the hypothesis of archaic admixture certainly presents itself as a very attractive one, given the data we have at this time.

Bonnie

andrew said...

While it is a close question, the case for this as a uniparental introgression, rather than a very early AMH sourced uniparental genetic legacy is strong than any other I've seen.

I wonder if the statistical indicators that the whole genome may contain a small percentage of archaic admixture in the recombining part of the autosomal genome based upon linkage disequalibrium is present in the NRY-DNA haplogroup A00 individuals. If it is, this would greatly strengthen the argument that this is a recent (perhaps 13 kya) archaic introgression.

shenandoah said...

Mike said, "...chimpanzees are our extant sister group, not our 'common ancestor'."

Good point, and thanks for clarifying that for me.

Bonnie said, "It's pretty technical and a subject of great controversy -- I hope others will comment on dating, but first of all it's done by using an estimated mutation rate for how often SNPs occur, and then counting the SNPs from a known point back through the phylogeny to the earliest branching point.

"The Chimp and Gorilla sequences were used largely to determine which was the ancestral state of each SNP, and which was derived."

I can understand why dating / chronology of the haplotypes is so controversial. For example, I gather that SNPs are basically just variations in the genome between different individuals. It's difficult to pinpoint which individuals might be aboriginal, or simply indigenous to particular regions. (Mike's point referring to possible morphological standards was well taken.)

Unknown said...

What were the 95% Confidence limits for the higher mutation rate ? If it was anything like what the limits are when using the lower mutation rates that they are reporting here, -30% for the lower bound and +72% for the upper bound from the central estimate, then the 95% CI for the higher mutation rate at the CT node would be roughly 27,000 - 67,000. Which leaves plenty of wiggle room for a post toba OoA, and the earliest unambiguous recovery of evidence of modern humans out of Africa, which is really no earlier than 50 KYA.

Dienekes said...

Which leaves plenty of wiggle room for a post toba OoA, and the earliest unambiguous recovery of evidence of modern humans out of Africa, which is really no earlier than 50 KYA.

There is nothing ambiguous about the presence of anatomically modern humans in the Levant at around 100 thousand years ago.

Unknown said...

There is nothing ambiguous about the presence of anatomically modern humans in the Levant at around 100 thousand years ago.

Sure there is, whether or not there was population continuity from those specimens is largely ambiguous, whether or not they were physically modern human proper is kinda ambiguous and whether or not they were behaviourally modern is also ambiguous.

kalupitero said...

I need clarification on the mutation rate, please.

The study used the Kong rate, as Dienekes noted, and they clarified that it was between 4.39x10^-9 and 7.07x10^-9 [per year]. This is a rate of 1.10 to 1.80 per 25-year generation.

But the Kong study says "an average father’s age of 29.7, the mutation rate is 1.20 × 10−8 per generation." [see Dienekes link to Kong study] This would work out to 1.00 per 25-year generation.

Can someone explain the difference?

Dienekes said...

Sure there is, whether or not there was population continuity from those specimens is largely ambiguous, whether or not they were physically modern human proper is kinda ambiguous and whether or not they were behaviourally modern is also ambiguous.

There is nothing ambiguous about the modern human attribution of skulls like Qafzeh 9. "Behaviorally modern" is irrelevant, since there were no such people at the time. As for their continuity with subsequent humans, I see no compelling argument against it, except if one holds to a late post-50kya OoA. If one were to apply the same criteria to other parts of the world (e.g., East Africa), they'd find no more evidence of continuity, and indeed evidence against it (e.g., Herto being more archaic than Omo I despite being ~50ky younger).

Blender said...

@kalupitero,
There seems to be a typo in the sentence that Dienekes copied from the paper. A few lines above, the value reported is between 4.39x10^-10 and 7.07x10^-10 per year. This latter value agrees with the supplementary material, and fits the values used in the calculation.

Unknown said...

If one were to apply the same criteria to other parts of the world (e.g., East Africa),

Africa, most especially East Africa, is different, you can not apply the same criteria to it like you would Eurasia, that is to say the World (x Africa), which has primarily been a sink for human migrations and not a source, so it is easier to look for distinction in the homo varieties that have left the African homeland (on several occasions) and since then roamed Eurasia, however Africa is the opposite, the yin to Eurasia's yang if you will, as it has always been a source for humans and their predecessors, therefore I'd imagine it much more complicated in trying to find distinctions between human, pre-humans, pre-pre humans and so forth, basically distinctions will be blurred in Africa because it is the source and not the sink, so a different approach or paradigm needs to be taken when it comes to Africa.

terryt said...

"First, it forces us to consider the possibility of either (i) archaic admixture in Africa, or (ii) a much more ancient time depth of modern humans than the first fossils from Ethiopia dated to about ~200 thousand years ago".

There is a third possibilty. The haplogroup does indeed lie at the base of the modern Y-DNA tree, but the individual in which it first appeared was not 'modern' in any realisic sense. After all modern humans must have sprung from some earlier species. It is not as though a small group woke up one morning and suddenly discovered they were a different species from their neighbours.

"an African American Y chromosome that carries the ancestral state of all SNPs that defined the basal portion of the Y chromosome phylogenetic tree".

It does lie at the base, so even if introgression is the case that introgression was from a not too distantly related species.

"But, who were the people in Africa responsible for the introgression of A00 chromosomes?"

As I said, it may not represent 'introgression'. It may indeed lie at the root of the modern human Y-DNA. The origin of that Y-DNA may well be in West Africa. As you commented in relation to extra-Africa haplogroups:

"And, while there is no a priori reason for Y chromosomes and mtDNA to behave similarly"

I think it most unreasonable to suppose that they behaved the same.

"the extreme age and rarity of the A00 lineage point to the possibility of a highly structured ancestral population"

I have no doubt at all that the human population of Africa has been highly structured, probably for all of its existence from the time of Australopithecus.

"e.g., Herto being more archaic than Omo I despite being ~50ky younger"

Demonstrating that the African population was considerably diverse even 50,000 years ago.

"It seems also quite possible that anatomically modern humans are just a lot older than we think"

But there is no need to assume that the A00 mutation first arose in an individual who was already 'modern'.

Ted Kandell said...

Dienekes and all, just to set the record straight:

I did first identify, raise the funds for the Walk on the Y, and help pay for the sequencing and STR testing of these samples from the Perry family. (The family was named in the paper.) Permission by me was not granted to use this in a research paper by a for-profit corporation and my contribution was not acknowledged.

From: Ted Kandell

Subject: [DNA] Kit N64496 HAS NO HAPLOGROUP AT ALL - Help PAY for a WOTY andfull STR testing!

Date: Sat, 25 Feb 2012 23:55:36 -0800 (PST)


From: Ted Kandell

Subject: [DNA] Kits N64496 Perry and 215865 - Crossing the Middle-Passage:The "Holy Grail" of "Genetic Genealogy"

Date: Mon, 27 Feb 2012 22:46:08 -0800 (PST)


From: Ted Kandell

Subject: [DNA] Humans Out of Africa - or not?!

Date: Tue, 28 Feb 2012 10:52:46 -0800 (PST)


From: Ted Kandell

Subject: [DNA] "Y0" - A NEW ROOT FOR THE ENTIRE Y DNA TREE - Resequencing results for kit 215865

Date: Mon, 30 Apr 2012 04:19:37 -0700 (PDT)


From: Ted Kandell

Subject: [DNA] Unethical behavior by Family Tree DNA surrounding the Y0 discovery of the root of Y the tree

Date: Mon, 30 Apr 2012 03:57:55 -0700 (PDT)


From: Ted Kandell

Subject: [DNA] Y0 - unjustly withholding test results for kit 215865

Date: Mon, 30 Apr 2012 04:30:33 -0700 (PDT)


From: Ted Kandell

Subject: [DNA] Y0 Kit # 215865 - FTDNA - FIX IT NOW!!!

Date: Mon, 30 Apr 2012 03:59:17 -0700 (PDT)


NOTE: My thinking was not in accord with the thinking of Thomas Krahn, and he could not speak for me.

From: Nancy Grossman

Subject: [DNA] New proof of an African origin of Anatomically Modern Humans?

Date: Tue, 1 May 2012 05:22:32 -0500


From: argiedude
Subject: [DNA] A0* (y-dna) haplotype chart... and a dilemma involving DYS635

Date: Tue, 1 May 2012 20:02:29 -0300



From: Thomas Krahn

Subject: Re: [DNA] "Y0" - A NEW ROOT FOR THE ENTIRE Y DNA TREE- Resequencing results for kit 215865

Date: Mon, 30 Apr 2012 21:07:43 -0500

____________

kalupitero said...

@blender

That would make the difference much worse, it would make this study's mutation rate 0.11 to 0.18 per 25-year generation.

Let's forget the 25-year conversion I made, that was a stupid complication. In this study, the mutation rate is 4.39 to 7.07 x 10^-9 per year. In the Kong study it is 4.04 x 10^-9 per year (see Dienekes link in this same thread to Kong study where the mutation rate is described). But the study says they used the Kong mutation rate, so I don't know why they describe it as 4.39 to 7.07.

Alvah said...

Dienekes said...
Sure there is, whether or not there was population continuity from those specimens is largely ambiguous, whether or not they were physically modern human proper is kinda ambiguous and whether or not they were behaviourally modern is also ambiguous.

Some would beg to differ both anatomically and behaviorally…
As to whether ‘transitional forms’ where in the Levant and/or Africa many cast doubt, for example; Corruccini.
Robert S. Corruccini, 1992. Metrical Reconsideration of the Skhul IV and IX and Border Cave 1 Crania in the Context of Modern Human Origins; AJPA April
“Skhul IV and IX, meanwhile, …join the cluster of supposedly early African/Mideastern AMHS (*my “pre-sapiens”), and then the above-described, increasingly heterogeneous Neandertal grouping. The Upper Paleolithic true AMHS exclusively cluster with one another, relatively far removed from these other groupings. Thus, the picture is one of overriding affinity among all the crania earlier than the European Upper Paleolithic, whether they be considered Classic Neandertal, Progressive Neandertal, AMHS, presapiens, or whatever.”…

”Continued facile reference to Skhul and Qafzeh as craniometrically “fully anatomically modern” is not responsible to the craniometric data (pg. 437).”

“The Border Cave cranium, so central to the course of “out-of-Africa” thinking despite its uncertain age, can support no special relationship to living African Homo sapiens.(pg. 441)”
“Qafzeh 6 as well as Skhul IV and V are well separated from later European AMHS. This calls into question blithe assumptions that Skhul and Qafzeh are, cranially, anatomical moderns.” (pg. 441)

As for biobehavioral modern non-sapiens may I quote Harvard’s David Pilbeam …

Pilbeam, David; In Major topics in Primate and Human Evolution, editors Wood, Martin, and Andrews, published by Cambridge (Cambridgeshire) New York 1986. pg. 335.

“Whatever the exact nature of the behavioral differences between modern humans and their ancestors, and of the transition between them there is a plausible case to be made for the argument that the biobehavioral gap was wide, that ‘archaic’ human behavior was different from the behavior of anatomically modern groups, and that we see in the ‘archaic’ the final representatives of a very long phase of human evolution, during which only limited changes took place… A case can be made that the nomen H. sapiens should apply only to hominids for which modern behavior patterns can reasonably be inferred: another name would then be needed for ‘archaic’ H. sapiens [Homo erectus].”

Dienekes said...

Dienekes said...
Sure there is, whether or not there was population continuity from those specimens is largely ambiguous, whether or not they were physically modern human proper is kinda ambiguous and whether or not they were behaviourally modern is also ambiguous.


I didn't say that, make sure you quote the right person.

Both Qafzeh 9 and Skhul 5 fall within the range of modern humans, and indeed the former is in the middle thereof

http://dienekes.blogspot.com/2011/09/12-6ka-humans-with-archaic-features.html

German Dziebel said...

"Both Qafzeh 9 and Skhul 5 fall within the range of modern humans, and indeed the former is in the middle thereof"

Qafzeh 9 is right next to a UP European but is 50-60,000 years older. It's unlikely that the skull didn't change for that amount of time. For comparison, Hofmeyr skull from South Africa is 36,000 years old and also clusters with UP Europeans but not with later Africans. This seems right. I would ignore Qafzeh 9, under the current date of 90-100,000 as noise in the data. Unlike other Skhul/Qafzeh skulls, Qafzeh 9 doesn't show brow ridges, hence it's likely a find introgressed from younger strata.

Skhul 5 has Neandertal-like brow ridges. No UP European specimen has it. One could weigh traits to lessen the significance of brow ridges in Skhul 5 and get it barely over the line into the modern human cluster, but the inconvenient fact remains.

So, Qafzeh 9 doesn't fit with modern humans chronologically and Skhul 5 morphologically. Neither of them fit the bill behaviorally, as they are found in association with Mousterian technologies.

Alvah said...

Dienekes, Please forgive my misappropriation of the quote, I should have known better than to link it to you. Here is another blast from the past, one that is also not out of date.

From: Howell, F.C., 1984, "Preface" to The Origins of Modern Humans: A World Survey of the Fossil Evidence, Eds. Smith FH, F. Spencer, New York: Alan R. Liss, Inc. 1984

"There is now a near consensus among students of human evolutionary biology that the origins of our own species, Homo sapiens, is somehow intimately linked with the first intercontinental ancient hominid, Homo erectus. However, neither the transformation of erectus to sapiens nor the transformation of ancient (archaic) populations of Homo sapiens to their anatomically modern successors (H s sapiens) are matters of agreement in this scientific fraternity. Undoubtedly, there are many factors that make this the case, and any reader of this volume will discern some of those that are most obvious. In fact, there is no consensus among the authors represented in this volume, although the major issues are generally well delineated, and the limitations of the diverse and often disparate lines of evidence are usually apparent (p. xiii.)."

Are alternatives to H. erectus available?