The latest edition of The American Journal of Human Genetics has two papers using “old fashioned” uniparental markers to trace human migration out of Africa and Siberia respectively. I say old fashioned because the peak novelty of these techniques was around 10 years ago, before dense autosomal SNP marker analyses, let alone whole genome sequencing. But mtDNA, passed down the maternal line, and Y chromosomes, passed from father to son, are still useful. Prosaically they’re useful because the data sets are now so large for these sets of markers after nearly 20 years of surveying populations. More technically because these two regions of the genome do not recombine they lend themselves to excellent representation as a tree phylogeny. Finally, mtDNA in particular is particularly amenable to estimates via molecular clock methodologies (it has a region with a higher mutational rate, so you can sample a larger range of variation over a given number of base pairs; you can use STRs, which mutate rapidly, for Y chromosomes, but there seems to be a lot of controversy in dating).
The papers are The Arabian Cradle: Mitochondrial Relicts of the First Steps along the Southern Route out of Africa and Mitochondrial DNA and Y Chromosome Variation Provides Evidence for a Recent Common Ancestry between Native Americans and Indigenous Altaians. Dienekes has already commented on the first paper. I am not going to take a detailed position on either, but I have to add that we need to be very careful of extrapolating from maternal or paternal lineages, and, assuming that population turn over is low enough that we can make phylogeographic inferences about the past from the present. For example, if you look at mtDNA South Asians as a whole strongly cluster with East Asians and not Europeans, while if you look at Y chromosomes you see the reverse. The whole genome gives a more mixed picture. Additionally, ancient DNA analyses in Northern Eurasia are showing strong discontinuities between past and present populations. So coalescence back to last common ancestor between two different lineages in two different regions may actually be due to diversity in a common source population more recently, which entered into demographic expansion and replaced other groups.
If you need the papers, email me. Some of you know the alphabet soup of haplogroups better than I do. Below are two figures which I think give the top line results.
Razib Khan’s degrees are in biochemistry and biology. He has blogged about genetics since 2002, previously worked in software development, is an Unz Foundation Junior Fellow and lives in the western US. He loves habaneros.
January 27th, 2012 at 7:41 pm
As I think you noted the problem is that where the genes are today may not clearly tell you who lived where in the past and that thinking it does can lead to staggering errors. Plus you only draw blood from survivors who may not be all that closely related to population that lived in a given region in the past. Evidence of major population replacements among modern humans is starting to stack up.
January 28th, 2012 at 8:33 pm
An interesting methodological feature of Reich, D., et al. “Reconstructing Indian Population History” (doi:10.1038/nature08365) was the following:
“The idea of fitting allele frequency differentiation to historical models was first explored by Cavalli-Sforza and Edwards [35], and here we extend it to trees with mixture. This approach contrasts with the STRUCTURE algorithm, which fits data without a tree [36], or a tree in which many groups split simultaneously from an ancestral population followed by mixture [37]. Although STRUCTURE is accurate for estimating individual mixture proportions in recently mixed groups, it is not clear whether its estimates of ancient mixture are biased because it does not model hierarchical relationships among groups, which could lead to inaccurate estimates of allele frequencies in ancestral populations. In contrast, we use a more realistic tree model, and provide a test of fit.”
Supplemental notes 3 and 4 to the study flesh out the idea. Phylogenetic trees yield demonstrable phylogenetic inferences it would seem to me. My 2c.
January 29th, 2012 at 3:31 pm
It is hard to tell what is going on in the Native American paper.
Previous Y-DNA and mtDNA studies have shown that there are some Y-DNA phylogenies (e.g. Q-M3 and its descendants) and some major mtDNA haplogroups (A2, B2, C1b, C1c, C1d, and D1) that are very old and found throughout the Americas.
There are other Y-DNA and mtDNA haplogroups that are found predominantly only in certain North American groups (Y-DNA Q-M242 (xM3), R1 -M127, C3b, mtDNA X2a, D2a1, A2a. A2b) that seem to be traceable to later migrations (the case of R1-M127 being absent from the first wave or part of a serial founder effect in early first wave Native American migrations is less than clear). Na-Dene and Circumpolar populations tend to have lots of atypicality, Northeast North American populations are rich inY-DNA R1 -M127 and mtDNA X2a but not necessarily some of the other atypical New World Y-DNA haplogroups like Q-M242 (xQ-M3) and C3b that are pretty specific to the Na-Dene and Circumpolar populations. The indigeneous American form of R1 sometimes called R1* splits off before the R1a v. R1b break and is also associated with Indonesia and the Phillipines (arguing for its presence in the first wave indigneous population followed by serial founder effects as that population spreads out, as does the presence of R1* in indigeneous people of South America).
A link between the Na-Dene and Circumpolar Q-M242(xQ-M3) populations and South Altaian populations at a recent date would make sense and could plausibly be refined by an improved South Altaian phylogeny of Y-DNA haplogroup Q (Q-M3 is basically a private American haplogroup so new phylogenies in Siberia should have much of an impact on it to show linkage). A link between all Native Americans via Q-M3 and South Altaian populations makes less no sense at pedigree mutation rates which tend to be more accurately calibrated to historical dates (although the whole Y-DNA dating scheme is fundamentally problematic) but could fit the evoluationary dates.
February 1st, 2012 at 7:00 am
Hello ohwilleke,
Could you please clarify your previous comment?
Thanks.