Analysis of multiple alleles reveals human family tree.

Reconstruction of Human Evolutionary Tree Using Polymorphic Autosomal Microsatellites

ABSTRACT Allelic frequencies of 182 tri- and tetraautosomal microsatellites were used to examine phylogenetic relationships among 19 extant human populations. In particular, because the languages of the Basques and Hunza Burusho have been suggested to have an ancient relationship, this study sought to explore the genetic relationship between these two major language isolate populations and to compare them with other human populations. The work presented here shows that the microsatellite allelic diversity and the number of unique alleles were highest in sub-Saharan Africans. Neighbor joining trees based on genetic distances and principal component analyses separated populations from different continents, and are consistent with an African origin for modern humans. For the first time, with biparentally transmitted markers, the microsatellite tree also shows that the San are the first branch of the human tree before
the branch leading to all other Africans. In contrast to an earlier study, these results provided no evidence of a genetic relationship among the two language isolate groups. Genetic relationships, as ascertained by these microsatellites, are dictated primarily by geographic proximity rather than by remote linguistic origin.

Fig. 3. Principal components (PC) analyses based on allele frequencies of 182 microsatellite markers in 19 worldwide human populations. A: Two-dimensional plot of PC1 vs. PC2, including chimpanzee samples. Bidimensional plots of PC1 vs. PC2 (B), PC1 vs. PC3 (C), and PC2 vs. PC3 (D), excluding chimpanzee samples.

You’ll need to enlarge this one to read it.

As can be seen from this analysis of multiple loci (182 to be precise) Greeks are neatly placed in with the rest of Europe, effectively contradicting the dumbass ‘study’ by a Macedonian from a few years ago that tried to claim they were all originally from Ethiopia and not closely related to other Europeans.

This paper supports the out of Africa theory, and I have a quibble with it. The number of unique alleles in Africa is 125, but in the rest of the world is 268, showing more unique DNA ‘out of Africa’ than in it. This would seem to contradict the OOA theory somewhat. I mean, Europe has had a much smaller human occupation period than Africa, and smaller population, but the ‘unique allele’ number is only about half of Africa’s (125 v 59). Bearing in mind that humans have been in Africa in the the longest and probably had the largest modern human population for most of our evolution, …  there should be a lot more difference between the two. A lot of ‘unique alleles’ seem to appear in too short a time in the non-African population.

5 responses to “Analysis of multiple alleles reveals human family tree.

  1. This paper supports the out of Africa theory, and I have a quibble with it. The number of unique alleles in Africa is 125, but in the rest of the world is 268, showing more unique DNA ‘out of Africa’ than in it. This would seem to contradict the OOA theory somewhat.

    Notice please that the African sample is much smaller than the non-African one, what may explain that very well. Also non-Africans (specially Asians) are many more people than Africans and have gone by more separate process of divergence: genetic structure is much more clear in Eurasia, Sahul and America than in Africa, where remixing has been going on all the time.

    Europeans are like 500 million and Africans like 800 million people (vey roughly the same), yet Europeans have half the diversity of Africans. Asia has like 4 or 6 times the population of Africa and yet it shows not more than double diversity. The combination of sample size and representativity (Europeans are best represented – so typical) and actual demographics would seem to explain that apaprent contradiction, right?

    Personally I find more contradictry that North Italians are closest to North Europeans in the NJ tree, while they appear somewhat apart from the main European (and South Asian!) cluster in the PC graphs. But those are the “mysteries” of statitical analysis and it is very likely that a different study, with slightly different samples and “weights”, will yield also different results. The main conclussions, like those supporting OOA (San separating first and so on) are consistent though in all studies, be them of aDNA, Y-DNA, mt-DNA and even older studies on proteinic markers and avant-guard ones on X-DNA. OOA seems impossible to refute from a genetic perspective, sincerely.

  2. Yes, but Europeans have been in Europe for a very short span of time compared to the African population’s in situ time; the earlier population barely made it into Northern Europe, then it was cleared out by the ice age, and repopulated when the ice age finished, so your looking at a total occupation time of 25k or so,with a pretty low population most of that time. The UK has only had a human population in it for about 17k total. This makes a level of uniqueness half that of Africas still pretty high (4.5%). The Asian population was also pretty low due to ice ages too, until recently. You have to remember that Europes Neanderthal population is thought to have been tiny (10,000 top), so they wouldn’t have had a massive amount of ‘uniqueness’ to have contribute. They also they seem to have had a population bottle neck of their own, which would have reduced their own diversity.

    I’ve got studies on X the chromosome that directly contradict the OOA pradigm.
    http://www.pnas.org/content/96/6/3320.abstract
    http://mbe.oxfordjournals.org/cgi/content/abstract/22/2/189?rss=1
    http://www.genetics.org/cgi/content/abstract/178/1/427

    One European variant of red hair has an MCRA age of about 80k. There’s quite a few traces of ‘not fitting the paradigm’ DNA out there.

    I’ve just got too many DNA studies showing very ancient non African genes to go with the OOA. More a ‘weak Eden’ scenario, with low levels of introgression. You have to remember there were never that many Neaderthals, so any interbreeding would have been pretty limited.

  3. Pan-European diversity as such is not subject to North European prehistory. Apparently the most important province of European UP is by large the Franco-Cantabrian region, with other areas also being somewhat important, including the Rhin-Danub province (which probably was not 100% depopulated at the LGM anyhow).

    The Asian population was also pretty low due to ice ages too

    That claim doesn’t make much sense to me. Surely Northern Asia was affected seriously but I have no particular reason to think that this was the case with all the Southern Asian strip, from Turkey and Palestine to South China and Indonesia, that surely enjoyed a much more benign climate (LGM permafrost only reached as far south as Beijing). Even today most of Asian population (and that means most of all human population) lives in that tropical/subtropical strip.

    IMO Eurasians must have experienced a major expansion after the OOA event, just because of finding so many more available niches (territory). Additionaly I am of the opinion that the OOA happened much earlier than normally considered, before the Toba event, probably c. 110,000 BP, what would give Eurasians a lot more time to diversify on their own.

    I’ve got studies on X the chromosome that directly contradict the OOA pradigm.

    I think I’ve read all them and there is nothing sufficiently clear. The anomalous clade is found in Africa too and there is a problem of major undersampling for Africa and zero sampling for India (the most central region re. Eurasian dispersal).

    Additionally X-DNA research is still in its infancy and, as happened with mtDNA and Y-DNA before a lot more may and will probably be learnt before our conclussions can become solid. Remember, for instance, the so many misclassified haplogroups in Y-DNA studies in the late 90s and even early 2000s (there was for instance a “rare” Hg21 found almost only in Basques that now has been relocated pretty much down inside R1b1b, while Hg1, the conceptual precursor of R1b, happened to be several clades instead). As X-DNA research seems to be advancing much more slowly than other autosomal research did in the past, I suspect that we will not be able to draw solid conclussions on it before a decade in the future or so.

    One European variant of red hair has an MCRA age of about 80k.

    I am very skeptical of TRMC age estimates. Compare for instance the wild ages (proper of H. erectus) given to Lactase tolerance clades in your other post (from an old study) and the modern understanding that these are actually Neolithic or something like that. But there are many other examples. TRMC age calculation is just way too dependent on variables that are themselves not well defined (mutation rate is certainly one but effective population size is always a wild guess, even more than the previous) and are filled speculatively. This is much like ancient mythologies and even presumpt serious works (like Marco Polo’s book or the Catalan Atlas) placed the most strange monsters and peoples to live in those parts of the World that had not been explored.

    Many people and the press often too tend to treat such estimates as “God’s word” but they are nothing but educated guesses hidden behind an array of esotherical formulas. If serious, they can be a reference for more holistic analysis but they are never proof of anything on their own.

  4. I don’t remember seeing an age of other thena about 8k for lactase persistance. Where exactly?

    Some of the more recent TMRCA have been shown to be reasonably accurate from DNA studies on ancient European bones (some german ones about 10k old didn’t have the light skin allele, and other tests for lactose tolerance have shown limited distribution in the Neolithic). It’s not just X chromosome, there are other ancient genes that just don’t fit the OOA.

    We’ll all be bickering until they’ve finished sequencing the whole lot. Although that’s not without it’s own issues.
    http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.0030175

  5. Nabril Sabri Enatah et al, 2007. “Evidence of Still-Ongoing Convergence Evolution of the Lactase Persistence T-13910 Alleles in Humans” (doi:10.1086/520705):

    To further address the issue of the historical origin of the common LP mutation in two diverse populations—Finns and Fulanis—we first estimated the most recent common ancestor (TMRCA) of the LP H98 T-13910 alleles in the Finns, using LD-decay analysis for marker D2S3014, which shows the highest LD with the LP phenotype in the Finns.3 NS Enattah, T Sahi, E Savilahti, JD Terwilliger, L Peltonen and I Jarvelä, Identification of a variant associated with adult-type hypolactasia, Nat Genet 30 (2002), pp. 233–237. View Record in Scopus | Cited By in Scopus (198)3 Using a generation time of 25 years and the algorithm by Risch et al.,[23] and [24] we found an age estimate of 5,275 years (95% CI 4,875–5,640) for the Finnish alleles. Use of the same marker, D2S3014, in the Fulani Sudanese population in the LD-decay analysis gave an age estimate of 6,475 years (95% CI 5,875–7,100). With three flanking markers (D2S3013, D2S3015, and D2S3016) that show less LD in LP alleles,3 the average square distance (ASD) method that used the Ytime program19 gave an age estimate of 9,252 years (95% CI 100–34,000) in this population (table 6).

    Some of the more recent TMRCA have been shown to be reasonably accurate from DNA studies on ancient European bones (some german ones about 10k old didn’t have the light skin allele, and other tests for lactose tolerance have shown limited distribution in the Neolithic). It’s not just X chromosome, there are other ancient genes that just don’t fit the OOA.

    I could not argue that some TRMC age estimates are not really good hunches. But I find that too often they appear to contradict each other and even the logic that archaeology seems to provide (for instance it’s obvious that Y-DNA E-V13 must have spread with Neolithic, because of its geographical distribution in Europe, yet you can see some people arguing it is from the Bronze Age because of their particular TRMCA calculations). So I look at them as just a reference, not as evidence. I think it is what they are actually worth: they are elaborate estimates, not unquestionable hard data.

    Also I have some reasons to think that the Human-Chimp divergence timeline on which much of the MC hypothesis is founded is normally underestimated. Check this post at my blog, where I ponder that all good TRMCA estimates should probably be expanded by at least 15% (on later though, 30% is probably more realistic).

    We’ll all be bickering until they’ve finished sequencing the whole lot. Although that’s not without it’s own issues.

    Sure. It is impossible to prove beyond doubt, I understand, that the admixture between the two Homo species is zero. But I also find it somewhat annoying the attempts to “prove” it is non-zero without clear evidence. So far, if it’s non-zero, it seems minimal anyhow: otherwise we would have already stumbled upon some uncontroversial evidence, I think.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s