October 13, 2009

Phylogeny of mtDNA haplogroup M in India (Chandrasekar et al. 2009)

PLoS ONE doi:10.1371/journal.pone.0007447

Updating Phylogeny of Mitochondrial DNA Macrohaplogroup M in India: Dispersal of Modern Human in South Asian Corridor

Adimoolam Chandrasekar et al.

Abstract

To construct maternal phylogeny and prehistoric dispersals of modern human being in the Indian sub continent, a diverse subset of 641 complete mitochondrial DNA (mtDNA) genomes belonging to macrohaplogroup M was chosen from a total collection of 2,783 control-region sequences, sampled from 26 selected tribal populations of India. On the basis of complete mtDNA sequencing, we identified 12 new haplogroups - M53 to M64; redefined/ascertained and characterized haplogroups M2, M3, M4, M5, M6, M8′C′Z, M9, M10, M11, M12-G, D, M18, M30, M33, M35, M37, M38, M39, M40, M41, M43, M45 and M49, which were previously described by control and/or coding-region polymorphisms. Our results indicate that the mtDNA lineages reported in the present study (except East Asian lineages M8′C′Z, M9, M10, M11, M12-G, D ) are restricted to Indian region.The deep rooted lineages of macrohaplogroup ‘M’ suggest in-situ origin of these haplogroups in India. Most of these deep rooting lineages are represented by multiple ethnic/linguist groups of India. Hierarchical analysis of molecular variation (AMOVA) shows substantial subdivisions among the tribes of India (FST = 0.16164). The current Indian mtDNA gene pool was shaped by the initial settlers and was galvanized by minor events of gene flow from the east and west to the restricted zones. Northeast Indian mtDNA pool harbors region specific lineages, other Indian lineages and East Asian lineages. We also suggest the establishment of an East Asian gene in North East India through admixture rather than replacement.

Link

50 comments:

terryt said...

So now we have 64 M lineages, of which just a few managed to seep outside India, presumably very slowly. So much for India being a major Paleolithic migration route.

eurologist said...

That just shows M is ancient in India, and most subsequent migrations happened early.

Anonymous said...

This is a good study. I don't consider the origin point of the "Ms" in situ to be too significant. I doubt any of the female carriers of the "Ms" were like any modern humans today, and unlike any living Indians today. Mitochondrial haplogroup M and N are ancient, and the subcontinent of India, a part of Gondwanaland that slammed into the Eurasian continent, is a landmass conducive to human population growth, so India containing many subclades of mtDNA M is understandable. The migration route of AMHs out of Africa was via south India, and it proceeded to SE Asia, Australia, and then north to the east Asian coastlands. The subcontinent was not left unoccupied but acted as a conduit to further migration which has persisted into the modern era and influenced the cultures and religions of mainland SE Asian countries and islands.

The East Asian admixture, if that is what it is called considering it is ancient and probably precedes the formation of the East Asian racial groups, is really a no brainer conclusion. STP studies of Indians shows an East Asian or mongoloid inheritance of many of the northern South Asians.

Gioiello said...

We are racist for our theory. Indians are altruistic in demonstrating that mtDNA M is born in India and deep rooted there from at least 50ky.
But by demonstrating this, they demonstrate also that hg. R has nothing to do with ancient India. India, like America, is the result of Indian women and “European” men.

German Dziebel said...

The principal investigator, Adimoolam Chandrasekar, is the supporter of the back migration of YAP+ Y-DNA from Asia into Africa. See his "YAP insertion in South Asia" (2007). He's looking for the same signal in mtDNA. The established migration of some U and M lineages into North and East Africa doesn't match the magnitude and the antiquity of the YAP+ back migration. So Chandrasekar simply formulates the problem (which is an old problem dating back to at least Cann et al. 1987) as the lack of ancient M lineages in Africa and the lack of L3 lineages in India.

Maju said...

I'm still reading the paper, so no much to say yet on it. However, Gioello wrote:

But by demonstrating this, they demonstrate also that hg. R has nothing to do with ancient India. India, like America, is the result of Indian women and “European” men.

What does this rant mean if anything at all? Are you talking of mtDNA R (highest top level diversity in India very clearly) or Y-DNA R (more complex maybe but also with apparent older phylogeny in India)? Whatever the case you are probably at least partly wrong.

India (or rather South Asia) is a very large subcontinent with very old phylogenies. It is most likely that all or at least most European (and West Eurasian in general) lineages are one way or another derived from Indian ones. Guess that Y-DNA is less clearly cut, specially for IJ (and of course E), but there is virtually no argument to make regarding mtDNA in this aspect.

Maju said...

So now we have 64 M lineages, of which just a few managed to seep outside India, presumably very slowly. So much for India being a major Paleolithic migration route.

The paper argues that "East Asian" lineages M8 (that includes CZ), M11 and M12'G in India "are rather ancient with ages >50,000 years". These lineages are AFAIK also old in East Asia.

So much for your preconceived pseudo-scepticism.

Maju said...

From the paper:

It is apparent that all the ancient lineages under analysis emerge directly from the root of the macrohaplogroup M.

This is an important confirmation. As I have mentioned before, very few haplogroups at any phylogenetic level, approach such an immense starlike structure of more than 40 top-level sublineages. Only H does but not even this Western clade has so many top-level subclades (31 "only"). Then come, at quite some distance other macro-haplogroups like N and R (12 and 16 top tier sublineages respectively). For comparison L3 only has 7.

It is my understanding that this structure implies a rather sudden and dramatic demographic expansion as which can only happen when a population expands in virtually virgin, essentially uninhabited, land, from a small number. This M starlike structure is the most clear signal of the Eurasian demic expansion that happened after soon after the OOA (or maybe, if we like it better catastrophist, after the Toba supervolcano caused a massive bottleneck among early Eurasians). Of course the comparatively smaller N and R starlike structures do also represent that same expansion in one way of another. However the starlike structure of mtDNA H must represent some other similar event but circunscribed to West Eurasia, such as the colonization of Europe some 45-45,000 years ago (and there was also a supervolcano then, for the catastrophists).

Maju said...

From the paper, for TerryT in particular:

The North Asian route could not get support from mtDNA due to the lack of basal M, R, N lineages in northern Asians, thereby ruling out the existence of a northern Asian route. And follows a large list of reference footnotes.

Gioiello said...

Maju, I have no time to answer you deeply. My reasoning was simple: if mtDNA M is deeply rooted in India (50kyBP), if YDNA R has been originated in India, we should have mtDNA M everywhere we have YDNA R. This isn't, then YDNA R has nothing to do with INDIA. Is it clear, or not?

German Dziebel said...

I don't know, guys. I still see it as a "slippery slope" model: a few archaic lineages C/Z, D, M9-M12 entered India from the east when populations were still small (because they're found all the way into the Americas, M included), then a cornucopia of new ones emerged in situ as a result of rapid population expansion.
L3 is not found in India, hence it's probably never been there. And this study also confirms that there're no archaic M lineages in Africa.

Maju said...

My reasoning was simple: if mtDNA M is deeply rooted in India (50kyBP), if YDNA R has been originated in India, we should have mtDNA M everywhere we have YDNA R.

Why? South Asia also seems ancestral for other lineages, notably mtDNA R, which is relevant in the sense you mean, as almost all West Eurasian mtDNA is R. Similarly, South Asia also hosts other Y-DNA lineages like H, L or C5, which have not seen any significative expansion out of India.

Of West Eurasian mtDNA lineages all but maybe X and M1 have a marked relation with South Asia (sister lineage/s only found there) and can be argued to have originated in the subcontinent. This is very clear for the case of mtDNA R.

L3 is not found in India, hence it's probably never been there. And this study also confirms that there're no archaic M lineages in Africa.

L3 is in India and everywhere, in the derived form of M or N. However if you mean L3(xM,N), you are right. Though I don't know why it would matter, as some sort of "bottleneck" or rather founder effect is always considered when dealing with OOA (otherwise Eurasians would have all the African diversity and we don't). Still, 2/7 L3 subslineages are Eurasian, what means that the bottleneck was not so severe: just a rather wide founder effect.

In the long term you won't find L3-underived anywhere, the same you don't find M-underived, etc. All lineages evolve one way or another. Our Indian and otherwise Eurasian L3 is necessarily in the form of M and N, unless there is a third Eurasian lineage not found till now. If it makes you happier, just call M and N with their more meaningful names of L3m and L3n, ok?

Maju said...

Oh, man, that last comment was not from some ignorant incidental poster but German Dziebel, who is obsessed with his own theory and with not understanding genetics: he is not ignorant, he wants to remain that way.

Would I have noticed, I would not have replied because it's a waste of time and virtual ink. I think you are confusing people with your false ignorance: you know very well all that by now.

German Dziebel said...

Luis, no L lineages are found outside of Africa. Not L0, not L1, not L2, not L3. Nowhere. No M and N lineages are indigenous to Africa. This is not a founder effect or a bottleneck. It means that the current mtDNA phylogeny is wrong. Just do me a favor, reverse a few mutations and show me what happens.

"L3 is in India and everywhere, in the derived form of M or N."

I responded to this nonsense a few times already. According to you (and you owe this thought to some awkward phrasing found in certain papers), the whole world is Africa. You can build circular arguments but they're not convincing.

terryt said...

"The North Asian route could not get support from mtDNA due to the lack of basal M, R, N lineages in northern Asians, thereby ruling out the existence of a northern Asian route".

The lack of basal lineages in no way rules a northern Asia route. As you yourself later say, 'In the long term you won't find L3-underived anywhere, the same you don't find M-underived, etc. All lineages evolve one way or another'. So why would you expect basal lineages to survive? especially as we can expect selection, bottlenecks and drift to have been severe along such a route. And It's possible the authors are a little committed to the idea that India has been important in our origin story?

Leaving out M and R, which I'll concede are southern, we do actually find some very early N lineages through Central Asia: A, X and, in the Far East, Y/N9.

"South Asia also seems ancestral for other lineages, notably mtDNA R".

Why? I see Ian Logan still has R on the Y/N9 line, so it need not be Indian in origin. He also has New Guinea P as being close to R's origin. It's quite possible that R is SE Asian in origin.

Gioiello said...

If mtDNA R is very ancient, also Tuscany has it. A relative of mine, I tested by SMGF, is resulted R0a, the same haplogroup that the last paper on Etruscans (Arredi et alii if I remember well) found they haven’t in common with Middle Easterners and the authors probably too hastily thought that what was common was from there (I can demonstrate that it could be the contrary too).
See: Mitosearch W9E9S.

Gioiello said...

Of course the paper is “Alessandro Achilli et alii, Mitochondrial DNA Variation of Modern Tuscans Supports the Near Eastern Origin of Etruscans”.

Gioiello said...

The European R0a with no mutations in HVR2 (Mitosaerch WQYNW and QZQJ6) and the basic mutations of R0a in HVRI (126C, 362C), is probably the European R0a closer to CRS and probably European mtDNA H arose from a common European haplotype.
That Tuscany retains the descendants of these ancient haplotype demonstrates, I think, that probabbly Italy was the source, remembering also the probably mtDNA H of Paglicci (28kyBP).

Maju said...

I think we're all the time talking CRS as in HVS-I, not the full sequence. HVS-2 does not seem to be relevant as is almost never tested.

German Dziebel said...

Apropos Terry and Luis: I just read a recent paper on the distribution of grammatical forms, namely Nichols, "Why Are Stative-Active Languages Rare in Eurasia: A typological perspective on split-object marking" in The Typology of Semantic Alignment (2008). Nichols's approach called "population linguistics" is capable of identifying zones of language spread and language accretion. In a number of publications, this one being the latest, she noted the peculiar absence of ancestral grammatical forms (active-static, head-marking languages) in Northern Eurasia (!). They're widely spread in America, Oceania and in such pockets as Basques and the Caucasus, somewhat common in southern Eurasia (Sino-Tibetan) and Africa but are utterly missing in Northern Eurasia, including Indo-European, Uralic, etc.
She calls Northern Eurasia a vast "spread zone" meaning a territory of constant population movement and language shifts which obliterated those ancestral grammatical forms. Alternatively, America, Oceania, the Caucasus and the Basques are the main refugia preserving older linguistic forms.

Maju said...

The lack of basal lineages in no way rules a northern Asia route.

It's Chandrasekar's conclusions, not just mine.

As you yourself later say, 'In the long term you won't find L3-underived anywhere, the same you don't find M-underived, etc. All lineages evolve one way or another.

Apples and oranges. Basal lineages are indeed found in tropical Asia and mainland Oceania but in Northern Asia nor in Europe or America, they are all clearly derived of this southern Asian huge top level diversity. The case is very clear.

So why would you expect basal lineages to survive? -

Because we find them in the south and SE but not in the north or NW. Not underived as you claim (absurd!) but highly derived in fact in many many subbranches (depending on the particular lineage). Instead every single northern Eurasian lineage is derived, often several nodes downstream, from a southern one.

And It's possible the authors are a little committed to the idea that India has been important in our origin story?-

And is it possible that you are prejudiced against such idea? You are obsessed up to the point of resulting suspicious with India being driven out of the equation. I am beginning to suspect a racist motivation, conscious or not.

we do actually find some very early N lineages through Central Asia: A, X and, in the Far East, Y/N9.

X, as you know, has a clear West Asian origin. A is highly derived and the obvious product of a most accidental founder effect, N9 looks SE Asian in origin and spread northwards along the coast, as expected. There are 12 basal N lineages and you have what? Only one.

Of course, N does look more SE Asian than South Asian, where N basal lineages are somewhat rare (N2, R, N1'5) but that's not what you are interested in knowing, right? You want to argue your hypothesis until Hell freezes.

Well, Hell has already frozen (cf. Dante). Stop it, please.

Maju said...

...and Africa...

Right?!

However, Bickel and Nichols make elsewhere a much less conclusive case for such patterns to matter in fact (and Basque is described as not having that pattern in the very first paragraph, but being ergative instead). Maybe the case that is Bickel and not Nichols the leading researcher here puts the emphasis elsewhere. Slippery science this one of linguistics, indeed.

Gioiello said...

Maju says: “I think we're all the time talking CRS as in HVS-I, not the full sequence. HVS-2 does not seem to be relevant as is almost never tested”.

We can discuss if HVRII isn’t “relevant”. Anyway we have two mutations between European R0a and the source of H after tens of thousands years. Find you a R closer to H elsewhere.

terryt said...

"N does look more SE Asian than South Asian, where N basal lineages are somewhat rare (N2, R, N1'5)".

And are most easily explained as being immigrants to South Asia.

"every single northern Eurasian lineage is derived, often several nodes downstream, from a southern one".

What? Two contradictory statements. Unless you're prepared to concede that South Asia is not the source of these southern lineages. If india is not the source how did they get to SE Asia?

"A is highly derived and the obvious product of a most accidental founder effect".

Where from?

"X, as you know, has a clear West Asian origin".

And fairly well north in West Asia. I've even seen Beringia quoted as its origin.

"N9 looks SE Asian in origin and spread northwards along the coast, as expected".

Your theory.

Maju said...

If india is not the source how did they get to SE Asia? -

Flying in a UFO. *sarcasm*

IMO underived N migrated there with the earliest M ones. N must be younger than M, whatever MCH fans think. You already know I think that - why do you even ask?

Where from? -

As N (the only branched upstream node) should be from SE Asia, from there.

And fairly well north in West Asia.

X1 is more North African than Asian and certainly is not found out of West Eurasia.

I've even seen Beringia quoted as its origin.

Of all X?! Pick better your readings.

Not even X2 has enough basal diversity out of West Eurasia for that even to be a half-reasonably. Obviously the rare North American Native X2 variants are a derivation, the only one in mtDNA, of a West Eurasian lineage. However, it is true that the intermediate steps, excepting some Altaian X2 (of the Western type) are lost by now.

Your theory.

Naturally.

German Dziebel said...

I'd be curious to know how do you, Luis, root your trees? Do you use a primate sequence, a Neanderthal sequence, a Mungo man sequence, or a "nuclear insert" in mtDNA?

Reidla et al. (2003, 1181) published an interesting caveat regarding the phylogeny of X haplogroup: "Moreover, haplogroup X is subdivided into two major subhaplogroups, designated “X1” and “X2.” Subhaplogroup X1, represented by a single Druze mtDNA in figure 1, differs from the root of haplogroup X by eight coding and three control region transitions and lacks the two transitions (195 and 1719) that characterize X2. These two nucleotides are rather mutable (Finnila ̈ et al. 2001; Herrnstadt et al. 2002); thus, it cannot be completely ruled out that X1 is indeed a subset of X2 that reverted at both nucleotide positions."

I personally haven't seen an article in which X was given a Beringian origin but A, X and (Australian) S are closely related lineages, hence I doubt X emerged in West Asia, as A and S aren't found there. East Asia, Beringia or America make more sense as its places of origin.

Maju said...

So you think that X2 might not exist... Interesting.

I don't think it changes much, specially when X1 is defined by a single control region marker (which looks much more feeble to me in principle).

A, X and (Australian) S are closely related lineages.

Not anymore than anything else under N: they are just three of 12 known lineages that apparently hang from that node, including macro-haplogroup R.

But anyhow, FYI, I recall that recent study by Alexe et al., who researched all the mtDNA tree all over (using a different method than the usual one of SNPs: K-means analysis) and, while they recycled completely the Eurasian (and American, etc.) part of the tree, even scrapping macro-haplogroup R and other "heresies", the basic structure of the global tree stood their efforts: L1, L2 and L3 (including what we still call M and N) persisted even against such titanic and unconventional efforts, only the structure of M and N (always under L3) was altered.

Anyhow, I think that the paper had little merit - though it's always good to read something that challenges it all, even if it fails. You may enjoy reading it but it's not going to give you the reason on your out of America fantasy.

German Dziebel said...

"So you think that X2 might not exist... Interesting."

Actually it's X1.

Alexe uses primate sequences to root a tree - this has been criticized. What about the nuclear insert? Without properly rooting a tree, it's hard to say which mutations are basal and which ones are derived. However, phylogeographic observations can be suggestive of the directionality of mutations. If L1, L2, L3 are African-specific, while M and N are found everywhere, including North Africa, I'd interpret M and N mutations as ancestral, while L1-L3 mutations as derived. The diversity within these lineages doesn't really matter as larger population size and gene flow in Africa will make its lineages more diverse than the lineages of more isolated populations with smaller effective population sizes outside of Africa.
By making L1-L3 mutations look ancestral phylogenetically (which is evolution only on a computer screen), the out of Africa theory arrives at a phylogeographic stalemate, with no L lineages found outside of Africa and with no early offshoots of M and N detected in Africa.

Also, Alexe's Fig 3 shows the trifurcation of M, N and L3 and not the subordination of M and N to L3. This looks somewhat like the D-E situation in Y-DNA.

"out of America fantasy."

Out of Africa is a lie, and the bigger the lie the more people believe in it.

Maju said...

"So you think that X2 might not exist... Interesting."

Actually it's X1
.

From your quoted text: [X1] "lacks the two transitions (195 and 1719) that characterize X2. These two nucleotides are rather mutable"...

And if you check PhyloTree, it is clear that the transition at 1719 is the only coding region mutation defining
X2, not X1.

Anyhow, what would seem is that there are some reasons to consider the X1/X2 as at least somewhat doubtful. However X stands and so do the lineages under the X1/X2 nodes, which are essentially West Eurasian (with one or two American exceptions).

Alexe uses primate sequences to root a tree - this has been criticized.

By whom? AFAIK what has been criticized is the use of the supposed but hardly clear Homo-Pan split age for MCH calculations. It is clear that using Pan sequences to root the human phylogeny makes all sense, why wouldn't it?

Without properly rooting a tree, it's hard to say which mutations are basal and which ones are derived.

Seriously, I do not understand what you mean. It's as simple as all human sequences having the defining mutations of L1-6, except those that belong to L0. That makes the root to be between L0 and L1-6. The same method applies to all branches at any level.

I'd interpret M and N mutations as ancestral, while L1-L3 mutations as derived.

You can't because these are sets and subsets. L3 includes M but M does not include L3. M has all the defining mutations of L3 (plus others) but L3 has none of the defining mutations of M.

That's how you root a tree: the inclusive set is always an upstream clade, the distinct subset is always a downstream clade.

Do your homework, seriously.

German Dziebel said...

You describe a hypothesis but you don't prove it. Phylogeography seems to disprove your phylogeny, as you don't find L lineages next to M or N in the same geographic area or in the same population. The only place where they're found in the same area is North and Northeast Africa and here M and N are derived from non-African M and N and not from L.

The same concerns America: if America was peopled, you would expect to find R, M and N lineages next to A, X, B, C and D at decreasing frequencies from North to South or you would find only R, M and N lineages because A, X, B, C, and D are either rare or non-existent in geographic areas of Asia leading to America. In actuality, you have the opposite: rare and supposedly recent lineages with a narrow geographic range were brought into America, while more common, older and hence more widely distributed lineages are not found in America.

You and others postulate lineage loss between Africa and, say, India and between Siberia and America but this doesn't explain the special selectivity of these losses twice in human history.

This is exactly how a false theory looks: new facts are retrofitted into a pre-existing idea, namely that America was peopled late in human history, while Africa, with its short and dark tropical people, is the homeland of modern humans.

My theory has no discontinuities: mutations frequently found in Amerindians are original and have the largest geographic distribution. African mutations are derived and have the most restricted geographic distribution. This corresponds to the linguistic map very closely: 140 families in America vs. 20 in Africa.

Instead of arguing with me and showcasing your fundamentalist proclivities, try to devise an mtDNA tree that will illustrate the current distribution of linguistic families. Work with different
tree topologies, root trees using Neanderthal, Mungo man or nuclear insert sequences (nuclear insert is from Zischler et al. 1995), simulate bottlenecks on route to Africa, add gene flow, etc. Use X chromosome, with basal B006 outside of Africa, as a model for a mtDNA phylogeny.

Take 9 bp deletion as an example. This mutation is found in Asia-America, in India and in Africa (high frequencies in Pygmies). What assumptions about mutation rate and bottlenecks do we have to make in order to postulate the common origin of this mutation plus a few SNPs also shared between the sequencies carrying this indel in Asia-America and in Africa?

Maju said...

You describe a hypothesis but you don't prove it.

Man: I'm not writing any scientific paper, I'm not the one who has laid out all those theories. I'm just falling into my altruistic outspoken side and trying to help you understand how genetics works. If you don't like it, then it's up to you.

... you don't find L lineages next to M or N...

Whatever. I have told you already a zillion times that M and N are subclades of L3.

You just want to prove that Earth is flat. Fine: it must be an entertaining but unrewarding endeavour.

German Dziebel said...

I understand how genetics works, no need to worry. But if you want to help science, rather than simply consume some of its products, you can consider other alternatives. An alternative is already a contribution to science: regurgitation is not.

M and N are not subclades of L3 unless you prove your phylogeny. So far phylogeography disproves it, as well as linguistics, kinship and other cultural elements. Archaeology is neutral, as it can accommodate pretty much any scenario. There're millions of ways in which trees can be generated. There're different tree topologies fitting the same data. Work from X chromosome to Y chromosome to mtDNA to look for alternatives. I suggested to you a few venues to explore.

Maju said...

I have my own priorities, which obviously are not the same as yours.

M and N are not subclades of L3 unless you prove your phylogeny.

M and N are subclades of L3 because its phylogeny has been proven: all tested M and N samples have the L3 defining mutations. There is nothing to discuss unless you have evidence of the opposite.

German Dziebel said...

Maca-Meyer rooted the tree in a chimp sequence. This is too far in the past and too hypothetical. Plus L3, M and N, even in Maca-Meyer's interpretation, are parallel lineages. Try to root the tree in the Mungo man sequence. It was more divergent than any L lineages. Note also this: "African L3 haplotypes are characterized by the absence of L1/L2-specific polymorphisms at nucleotide positions 769, 1018, 3594, 4104, 7256, 7521, and 13650″ (Herrnstadt et al. 2002. Reduced-Median-Network Analysis of Complete Mitochondrial DNA Coding-Region Sequences for the Major African, Asian, and European Haplogroups. Am J Hum Gen 2002 70 (5).) Finally, Maca-Meyer's tree contains a bunch of homoplasies: 16278 occurs 5 times in different parts of the tree.

Again, this is a hypothesis which, in my opinion, doesn't withstand a test.

Maju said...

What's wrong with rooting using a chimpanzee sequence? Would you prefer it'd be an American animal like a sloth? It would not make any difference, as long as it is an outgroup.

In 2001, L3 was still a tag not applied to N and M, as the phylogeny was still being worked out. However you can clearly see in fig. 1 how L3d and L3b are much more closely related to M and N (and viceversa) than to any other lineage. This is precisely the point of Maca-Meyer.

I'm not sure exactly when the tag L3 began to be used for all the macro-lineage but it's clear that this paper estabilishes the basics of such nomenclature, even if it still doesn't use it.

You're not going to fool me with semantics/nomenclature but I wonder if you can fool yourself. Whatever the name you give to that clade parallel to L2 in fig. 1, it is clear that it is monophyletic. We normally call that L3 and is virtually "official" nomenclature now.

Finally, Maca-Meyer's tree contains a bunch of homoplasies: 16278 occurs 5 times in different parts of the tree.

It can be a recurrent mutation or most likely different mutations at the same locus. This is sadly not too unusual in mtDNA, specially at control area regions (as is the case of all 16... loci) but not enough to blurr the phylogeny. Remember that the mitochondiral DNA is much shorter than any chromosome's and that the control region is just a small fraction of it (however mutating frequently enough as to give some useful keys).

Even if you remove the HVS mutations, the phylogeny stands. I don't think that Hermann's quote changes anything either (after all it's quite obsolete data, from the times L0 and the minor Ls were not even described). Two of those locus define L3 right now (per PhyloTree again). Basically he was noticing, it seems to me mutations describing L3. Back then only L3, L2 and L1 were known I believe (if we don't count M and N), so finding those specificities for L3 surely helped reinforce the distinct personality of this macro-haplogroup.

The current, quite solid, tree was not built in a single paper: it took the accumulative work of many researchers along maybe 15 years or more.

German Dziebel said...

What's wrong with rooting using a chimpanzee sequence? Would you prefer it'd be an American animal like a sloth?

Neanderthal, Mungo man, the nuclear insert are better rooting tools. Without them you can't decide which state was ancestral. Some trees are left unrooted, other trees are rooted.

The rest of your prose is full of apologies. mtDNA studies always had "noise" in them despite the 20 years of research. I would know - I was there.

Maju said...

Neanderthal, Mungo man, the nuclear insert are better rooting tools.

Mungo Man is no valid root: he clearly belonged to H. sapiens. He's no outgroup.

Neanderthal could be valid but the sequence was not known until 2003 and, anyhow, being aDNA, known only in a few specimens, guess it still arises some caution.

No idea what's "the nuclear insert".

German Dziebel said...

The nuclear insert (just like X chromosome's basal B006, it's frequent in Amerindians, rare in Africans, with all others in-between) was used to root the Mungo man sequence.

Zischler, H., Geisert, H., von Haeseler, A., Paabo, S., 1995. A nuclear ‘fossil’ of the mitochondrial D-loop and the origin of modern humans. Nature 378, 489–492.

The Mungo man sequence was the most divergent, hence, by definition, it fell outside of the modern human mtDNA variation. Anything of this sort can be used as a rooting outgroup.

terryt said...

"The Mungo man sequence was the most divergent, hence, by definition, it fell outside of the modern human mtDNA variation".

And that's the really interesting part, not descnded from L. The fact it appeared in Australia and is clearly 'modern' human sugests that the surviving haplogroups spread through a pre-existing population. And the modern haplogroups may not have their ultimate origin in Africa. They may derive from haplogroups immigrant to Africa. Unlike German I'm quite prepared to accept the current interpretation of haplogroup evolution.

Maju said...

*shakes head*

Mungo Man's remains were ony tested once in what seems to be a very controversial case with not enough guarantees. The remains were given to native tribals (a practice that I strongly oppose in what regards to remains older than say, 1000 years) and will never again be available for comparative re-testing most probably.

Sadly, I don't think that Mungo Man can be used for anything other than speculation at this point.

I see now what the nuclear insert is. IMO, unless it can be shown that somehow the fragment has resisted recombination havoc and is consistent for all humankind, it seems to me that it will also cause many problems and confusion. The Chimpanzee or maybe better the averaged Pan sequence seems the safest so far, though admittedly soon we will be able to contrast the modern human genome with a good quality array of several Neanderthal sequence, which should anyhow produce the same results.

Probably it has already been done, even if only informally, but I'm not aware of it. But otherwise, there would be for sure someone crying foul already.

German Dziebel said...

There's a theory by which the earliest Australians represent the first migration of anatomically modern humans out of Africa/ Levant (Mungo man's skull seems to cluster with Skhul/Qafzeh in some studies), while the rest of the world was peopled by the second migration (out of Africa, if you're a pygmy-hugger, or out of America if you're a serious scholar) of anatomically and behaviorally modern humans. The first migrants to Australia didn't survive, and the current Australian populations are the result of 2 separate migrations within the past 20,000 years. Hence, linguistically Australia is not as diverse as Papua New Guinea or America.

The nuclear insert is consistent for all mankind: it has all the synapomorphies of modern humans and none of those of Neandertals.

German Dziebel said...

Luis: "I don't think that Hermann's [read Herrnstadt's - GD] quote changes anything either (after all it's quite obsolete data, from the times L0 and the minor Ls were not even described)."

Luis, Herrnstadt is 2002, while Maca Meyer that you pointed me to is 2001. Which one is more obsolete?

Luis: "Mungo Man's remains were ony tested once in what seems to be a very controversial case with not enough guarantees."

As Kosmo put it the other day in a comment on a different article here, Mungo man results "have been essentially thrown out for no other reason than their wild divergence from expectation."

Well put.

Luis: "Sadly, I don't think that Mungo Man can be used for anything other than speculation at this point."

Everything that you write, including this last sentence, is a huge speculation. I hope you're aware of this.

Terry: "Unlike German I'm quite prepared to accept the current interpretation of haplogroup evolution."

Terry, X chromosome's basal lineage (B006) is most frequent in Amerindians, Y-DNA YAP+ lineages probably migrated into African from Asia and they account for the majority of African Y-DNA diversity.
The nuclear insert is found at high frequencies in America and at low frequencies in Africa, which according to Adcock et al. (the authors of the Mungo man study) implies that humans came to Africa, rather than originated there. African L lineages aren't found outside of Africa in either extant populations or in ancient remains. Non-African M and N lineages aren't indigenous to Africa. Ward et al 1991; Bandelt 2007 estimated that one single Amerindian tribe harbors mtDNA diversity worth of 80,000 years (sic!) of evolution.

All this seems to indicate very consistently that the conventional version of haplotype evolution is wrong. The first mtDNA study, by Johnson et al. 1983 as well as the subsequent study by Excoffier and Langaney 1987 identified out of Asia/out of America as a possible alternative to out of Africa.

Maju said...

I would have changed the word "obsolete" if I would have been able to edit. It's not really adequate. Anyhow, I only pointed you to MM because of vane hopes that you could maybe understand that, in spite of your extremely heterodox beliefs, M and N belong to L3, nothing else.

I wonder if this concept has already penetrated your mind or not.

I bet that when we cross paths again in the net you will again insist that M and N are not L3 and all that discourse that just can't hold.

As for Mungo Man. The tests were done apparently with little guarantees, for what I've read, and the fact that the remains have been "returned" to the Aboriginals, making any counter-test nearly impossible, makes us unable to discern the truthfulness or error in this oddball case.

Individual doubtful cases without a context have always been considered with some disregard and extreme caution. Otherwise we'd be talking vividly of the late Acheulean of NW Iberia, contemporary with Magdalenian and other oddball doubtful cases that are normally, when lacking further evidence, archived and disregarded until further evidence appears.

This practice is quite standard in prehistory and extends to many cases.

Anyhow, Mungo Man's apparent mtDNA was extreme but still within the range of modern humans. If you check figure 1 in Caramelli 2003, you can perfectly see that he is within modern human range, with modern day individuals being in some cases further away in the PCA.

Maybe more strange is to me that Cro-Magnon mtDNA is right in the middle of today's European range. Of course they are not that old.

Maju said...

And btw, I sincerely think that the comments sections in various blogs are not the place to post your caveats re. the mtDNA phylogeny. You should research the matter and publish something first, at least a series of blog pots, so the rest of the universe can follow your hypothesis consistently.

terryt said...

"the current Australian populations are the result of 2 separate migrations within the past 20,000 years. Hence, linguistically Australia is not as diverse as Papua New Guinea or America".

I'd agree with two (or more) migrations to Oz. However I'd put the first at more like 50k although I agree that many lines involved in that arrival have died out. A second migration seems to have arrived between 20k and 15k. The accepted reason for the lack of linguistic diversity through much of Oz (ignoring the far northwest) is that extreme drought exterminated the population through Central Oz but this region was later recolonised from within the continent.

"X chromosome's basal lineage (B006) is most frequent in Amerindians, Y-DNA YAP+ lineages probably migrated into African from Asia and they account for the majority of African Y-DNA diversity".

I agree completely that haplogroups don't tell the full story by any means. Individual genes move through individual species.

"Mungo Man's remains were ony tested once in what seems to be a very controversial case with not enough guarantees".

That is correct and so I agree we can't rely too much on the particular research.

German Dziebel said...

Luis: "Anyhow, Mungo Man's apparent mtDNA was extreme but still within the range of modern humans."

Check Adcock et al. 2001, Fig. 2A and you'll see that the range of variation in ancient Australian sequences overlapped with modern human sequences with LM3 falling outside of it. Nuclear insert overlaps with LM3 but is entirely outside of modern human variation.

"I sincerely think that the comments sections in various blogs are not the place to post your caveats re. the mtDNA phylogeny."

You started it by providing people with your grand theory of human haplotype dispersals. I just asked you about your rooting sequence. You didn't know what it was. I explained.

All my personal comments are relevant to the Chandrasekar paper under discussion, which again reiterated the lack of ancient M and N lineages in Africa and the lack of L3 lineages in India. This has implications for mtDNA phylogeny. I wrote a whole book and a post on anthropology.net outlining my views, so I think I did my job at this time. Now your job is to buy the book, educate yourself about kinship systems and linguistics and if you have any concerns you can write a post anywhere with a critique of the data presented. But your dogmatic belief in the infallibility of the out of Africa theory (mind you, you are not even a scientist, so what are your stakes?) deprives us of a minimal common denominator to interact without bickering.

I am planning to revise the mtDNA phylogeny to make it more consistent with phylogeography, demography, kinship systems and linguistics and to remove the "noise." However at this point I'm still waiting for a good genetics publication to provide me with a springboard.

"Anyhow, I only pointed you to MM because of vane hopes that you could maybe understand that, in spite of your extremely heterodox beliefs, M and N belong to L3, nothing else."

L3 is closer to M and N than either of them to L0, L1, L2, L5, etc. But being closer to doesn't mean they derive from L3. This is up in the air for me because of homoplasies, lack of the necessary mutations on the actually attested sequences, phylogeography, etc.

"As for Mungo Man. The tests were done apparently with little guarantees, for what I've read, and the fact that the remains have been "returned" to the Aboriginals, making any counter-test nearly impossible, makes us unable to discern the truthfulness or error in this oddball case."

This case looks "oddball" only because out of Africa was prematurely made into a measuring stick. There was nothing wrong with the Mungo man research. Yes, it's an unusual case of DNA extraction from a specimen of such antiquity and in such low latitudes (although the DNA from the Windover brain in Florida was also a case of remarkable preservation) but this doesn't make it invalid. Especially since it overlapped with an independently attested nuclear insert. It'd be good to replicate the results, sure, but if this is impossible I'd focus on trying to extract DNA from one of those African skulls to see how far in time L lineages go.

Maju said...

Now your job is to buy the book...

I'm not going to buy that book. I already know more than what I need to know about your kinship-based hypothesis and I know also that it is bad science. Maybe if it was fantasy genre...

L3 is closer to M and N than either of them to L0, L1, L2, L5, etc. But being closer to doesn't mean they derive from L3.

The branch that you seem to be calling "L3" in Maca-Meyer 2001 is actually L3b (or if you wish L3b'd'e'j), not all L3(xM,N). Only L3b and L3e (both within L3b in the most recent PhyloTree nomenclature) are analyzed in that study among "African" L3(xM,N)

Nomenclature changes and the least you could do is to check PhyloTree for the most actualized references. Do your homework, please (I'm not getting paid for these lessons).

German Dziebel said...

The fact that people call a clade to which M and N supposedly belongs "L3" is simply an out of African driven convention. M, N and L3x are sister clades with a tenuous connection to L1, L2 or L0. Think of D and E in Y-DNA as an example. Nobody calls Y-DNA D lineages E lineages. In the same way as D is at least as diverse as E or maybe more, L3 is as diverse or maybe less than M or N.

How can a person who isn't a scientist judge what science is and what is not? I have two doctorates and 20 years of experience across several science fields, including pop genetics. You are an amateur with huge gaps in knowledge, logic and ethics. At times your prose falls somewhere in the gray area between spam and cyberbullying.

Maju said...

Que no! M and N sister clade is not L3 as such but the various L3 sublineages as mentioned above, you are stubborn in this regard because you don't want to face reality. Do your homework and make some self-criticism.

For what I can read at Wikipedia-Scientific method, some guy named Charles Sanders Pierce, stated that the following are from worst to best the ways of finding the truth among humans:

1. The method of tenacity — persisting in that which one is inclined to think.
2. The method of authority — conformity to a source of ready-made beliefs.
3. The method of congruity or the a priori or the dilettante or "what is agreeable to reason" — leading to argumentation that gets finally nowhere.
4. The scientific method — whereby inquiry actually tests itself and criticizes, corrects, and improves itself.

You are stuck in method 1, you don't even care about method 2 and you are very far away from method 4 (science). You just have to be right no matter what.

It's boring, annoying and dismaying to discuss with you.

German Dziebel said...

At least, I made you read Peirce (note the correct spelling, smartie pants!) if not in the original. I have a whole chapter on Peirce in my book because he was one of the first students of kinship-based or relational logic in the 19th century.

"M and N sister clade is not L3 as such but the various L3 sublineages as mentioned above."

Very good. Now D lineages form an outgroup to all Asian and North-East African M lineages, while A, X, S and some other Eastern N lineages form an outgroup to the European-North African (Western) N lineages (including U). Some of the M and N lineages are found in Africa as well as outside of Africa, while all African "L3" lineages are restricted to Africa. African L3 sublineages are derived from certain lineages grouped under Western N submacrohaplogroup, U lineages included. I'm still working on the rest of African L lineages (L0, etc.) but most likely they're derived from X, B and some other "Eastern" lineages.

This is the tree topology that fits phylogeography, linguistics, kinship, etc. better. And it'll have fewer homoplasies than the current one.

Don't argue with it. It's a working hypothesis. I'll publish a revision at some point with all details in it.