Here is my analysis of the article “The Genetic Legacy of Paleolithic Homo sapiens sapiens in Extant Europeans: A Y Chromosome Perspective” by Semino et. al., Science *290 1155.
There is quite a lot of work published on this stuff and I am not intimately familiar with all (or most, or even a chunk) of the literature, so what I say may be outdated/wrong by now.
Anyway, this article is quite interesting. In it, they analyze many populations throughout Europe and the Middle East for their Y chromosome haplotypes.
Some background: The Y chromosome is passed from father to son. Unlike most other regions of the genome, most of the Y chromosome never exchanges genetic material with other chromosomes. Thus, it passes through generations untouched, except for the occasional mutation.
Mutations happen rarely. You can count the number of mutations which have arisen between two populations of males and determine a time when the two populations separated (as the mutation rate per locus is relatively constant). These collections of mutations are termed the “Y chromosome haplotype” (haplo because there is only one Y chromosome in each male cell).
Also, there are many different mutations out there, and some regions mutate faster than others (so there are different time rulers for different areas of the Y chromosome). By looking at the mutations, you can determine which Y chromosomes are related, and establish a phylogenetic tree to relate all of the Y chromosomes to each other.
For instance, take the word “dope” as a Y chromosome haplotype (established by looking at great ape populations and so forth). Arrows represent mutations.
dope -> hope -> hole -> holt -> molt -> moat -> goat
| | | | | |
v v v v v v
dole hype sole dolt malt moan
So anyway, Semino, et. al. went through Europe populations and found the most “pure” populations in certain regions. They then did cheek swabs to get DNA samples from men in these populations. They established a phylogenetic tree. No population has a pure haplotype, but many are dominated by one kind. This represents the so-called “founder effect” of the Y chromosomes, where the descendants of a few males (or perhaps one) came in and populated a region.
To the paper:
There are several haplotypes which dominate Europe. Two general families of haplotypes appear to have been in Europe since the Paleolithic era, by an estimation of the mutation rates and their subgroups. These are M170 and M173.
M173 is interesting – it has been in Europe for 35-40,000 years and extends into Asia and into the Americas. It has two subtypes which predominate in Europe. They appear with grossly contrasting geographic patterns – one (Eu18) appearing most frequently in Basques and declines eastward, the other (Eu19) appearing most frequently in the east and declining westward. The hypothesis presented is that M173 characterizes early Paleolithic inhabitors of Europe. During the last glacial maximum 22,000 years ago, Central Europe became uninhabitable. M173s retreated eastward and westward, and mutated at some point into Eu18 (in the Balkans) and Eu19 (in Iberia). After the glaciers melted, Eu19s and Eu18s repopulated central Europe. Eu18 actually spreads into India as well, perhaps marking the spread of Indo-European languages in the Yamnaia culture. Eu19 perhaps marks the Aurignac culture.
M170 has been for 20,000 years, and perhaps marks the “invasion” of a Middle Eastern population into Europe after the glaciers. They link this with the Gravettian culture. Gravettians and Aurignacs mingled, but remained largely isolated and non-interbreeding. One sees this if looking at other markers (mtDNA, which is passed mother to child and also doesn’t recombine with the genome).
Other haplotypes represent Neolithic Eastern invasion around 17,000 years ago. The frequency of these decrease from the Middle East into Europe, and represent demic farmers.
They then go on to correlate this to other data points. One is autosomal protein polymorphisms. They correlate these nicely to the Aurignac, Yamnaia, and demic farmer contribution. The other is mtDNA (mitochondrial DNA alluded to above). These mainly correlate, except some data points have no correlation between the two. Namely, mtDNA supports a 45,000 year old Middle East invasion which is not found in the Y chromosome stuff. Also, more of the mtDNA appears to be Paleolithic. This could be due to a number of factors – more rapid replacement of males than females during transitions (perhaps raping and pillaging), female migration (back to the OP), or perhaps other factors not explored. There is not much written on this – only a paragraph.
A nice bit of observation, but certainly not an earth-shattering result. Pretty cool what they can do with the genetics nowadays, though. There is a bunch published on this field, including some American Journal of Human Genetics papers that came out after this paper. I’ll try to read them in the next few days.