Population Substructure

How many SNPs to distinguish Japanese & Chinese?

razib — Fri, 05 Mar 2010 01:06:32 +0000

How many SNPs to distinguish Japanese & Chinese?

A new paper in PLoS, Rapid Assessment of Genetic Ancestry in Populations of Unknown Origin by Genome-Wide Genotyping of Pooled Samples:

Many association studies have been published looking for genetic variants contributing to a variety of human traits such as obesity, diabetes, and height. Because the frequency of genetic variants can differ across populations, it is important to have estimates of genetic ancestry in the individuals being studied. In this study, we were able to measure genetic ancestry in populations of mixed ancestry by genotyping pooled, rather than individual, DNA samples. This represents a rapid and inexpensive means for modeling genetic ancestry and thus could facilitate future association or population-genetic studies in populations of unknown ancestry for which whole-genome data do not already exist.

All fine & good. But I thought figure 4 was interesting. I've highlighted the part that I thought noteworthy:

The panel on the left shows that with 420 Ancestrally Informative Markers (AIMS) you can separate the Japanese and Chinese (from Beijing) pretty well. These are markers which exhibit a lot of between population variation. But note the last section: with 3,100 random SNPs you can generate the same separation. And remember that you'd need way fewer markers for African populations, since they have more diversity. In any case, I'll keep these numbers in mind when people ask me genetic distance related questions and I know that Fst numbers won't mean anything....

Citation: K. Chiang CW, Z. Gajdos ZK, Korn JM, Kuruvilla FG, Butler JL, et al. 2010 Rapid Assessment of Genetic Ancestry in Populations of Unknown Origin by Genome-Wide Genotyping of Pooled Samples. PLoS Genet 6(3): e1000866. doi:10.1371/journal.pgen.1000866

razib Thu, 03/04/2010 - 20:06

Tags

genetics

genomics

Population Genetics

population genomics

Population Substructure

I think that a length post summing up this kind of question, the difference between neutral markers (which Cavalli-Sforza tried to use) and significant difference (e.g. lactose intolerance), and the idea that groups are populations defined statistically by mixes of genes, rather than identities, and putting it in common sense language, would be valuable.

...with 3,100 random SNPs you can generate the same separation. And remember that you'd need way fewer markers for African populations, since they have more diversity.

Okay, here are my two questions Razib: If two bushmen have more diversity between them than a European and an East Asian, does that mean that the offspring of a Bushman and a European or an East Asian is less diverse at the genotype level than a full-blooded bushman? Also, I know that 'inbreeding' is not a healthy thing, but if anybody could do it with the least amount of problems would it be the bushmen?..

Razib,

Here's a theoretical analysis of resolving power as a function of FST, number of SNPs, etc.

http://infoproc.blogspot.com/2008/12/resolution-of-genetic-population.h…

With current technology any two European nationalities (or Chinese and Japanese) are easily distinguishable with random SNPs.

"For example, given a 100,000 marker array and a sample size of 1,000, then the BBP threshold for two equal subpopulations, each of size 500, is FST = .0001. An FST value of .001 will thus be trivial to detect. To put this into context, we note that a typical value of FST between human populations in Northern and Southern Europe is about .006 [15]. Thus, we predict: most large genetic datasets with human data will show some detectable population structure."

PS Thanks for that followup on lactose tolerance and height

Is there a reason why the AIMs seem to identify more Chinese individuals as Japanese than vice versa? Or is that just due to the sample size?

i wouldn't put too much stock in that, though i assume that the chinese population will be more diverse than the japanese (the main caveat being that the japanese are probably a recent admixture between yayoi, 3 parts, and jomon, 1 part).

What genes can't tell us about lactase persistence

razib — Wed, 17 Feb 2010 04:26:11 +0000

What genes can't tell us about lactase persistence

Lactase persistence results in the ability to break down the lactose sugar in milk as an adult, lactase being the enzyme which breaks down lactose. If one can not digest that sugar, and still consumes milk, then one exhibits the symptoms of lactose intolerance. Originally diagnosed as a disease it has come to light that 2/3 of the world's population is lactose intolerant, and, that this is probably the ancestral "wild type" (I would be interested if readers could name a mammal which exhibited lactase persistence aside from humans). Lactase persistence is a relatively new trait which emerged in the wake of cultural changes which were triggered by the Neolithic Revolution, primarily the domestication of cattle, goat, etc., and their utilization as a source of milk. The latter does not always entail from the former; the Gond tribe of India apparently has utilized cattle & goat for meat, but avoids drinking milk as adults (presumably they're lactose intolerant).

The ability to digest milk as an adult is a phenotype has been elucidated in terms of its genetics rather well in the past ten years, at least in relation to a host of other traits of interest such as height & weight. A few genes of large effect generate a dominant phenotype whereby lactase continues to be produced into adulthood, and, there are alternative genetic architectures which can lead to this persistence. In other words, there are different instances when mutations arose which conferred lactase persistence.

So as far as human genetics questions being illuminated in the "post-genomic era," lactase persistence is probably one of the successes. And yet there is much we don't know. A new paper addresses the issue by attempting to connect the dots in the holes of our knowledge, A worldwide correlation of lactase persistence phenotype and genotypes:

Results
We used surface interpolation of Old World lactase persistence genotype and phenotype frequency estimates obtained from all available literature and perform a comparison between predicted and observed trait frequencies in continuous space. By accommodating additional data on sample numbers and known false negative and false positive rates for the various lactase persistence phenotype tests (blood glucose and breath hydrogen), we also apply a Monte Carlo method to estimate the probability that known lactase persistence-associated allele frequencies can explain observed trait frequencies in different regions.
Conclusion
Lactase persistence genotype data is currently insufficient to explain lactase persistence phenotype frequency in much of western and southern Africa, southeastern Europe, the Middle East and parts of central and southern Asia. We suggest that further studies of genetic variation in these regions should reveal additional nucleotide variants that are associated with lactase persistence.

I've tried to look at the lactase persistence data myself, and one problem is how you measure the phenotype. This isn't like height or weight, where precision is pretty good. Their method attempted to taken into account the error rates in detecting lactose intolerance, both false positives and negatives. They also had to account for the fact that they had lactase persistence data on some populations, but no genotype data. But using geographical extrapolation methods they constructed a series of maps which show the clines of both the allele and trait frequencies. Here's one which illustrates the differences between the two sets of maps, with blue being the genotype over-predicting lactase persistence, and red under-predicting:

It is clear that there is a trend toward under-prediction to a much greater magnitude than over-prediction. This makes sense when you consider the possibility that there are likely other lactase persistence alleles out there which haven't been picked up in association studies because the samples aren't representative of the world's genetic variation.

Citation: BMC Evolutionary Biology 2010, 10:36doi:10.1186/1471-2148-10-36

razib Tue, 02/16/2010 - 23:26

Tags

evolution

genetics

Lactase Persistence

Population Substructure

evolution

Look, I love sucking on hind teat as much as the next guy but when are we gonna get your analysis of your readers? We need to know how many socially far-left, fiscally far-right fertile virgin Muslim PUAs lurk among us.

Or are you just too depressed to get to it as you note that (despite exhortations each annum) we're still not breeding.

Also relevant: How about another secular right poll. The goal would be to catch all of the Paulites and ban their IPs from ever accessing the page again.

Do you know if there is any evidence that people figured out the use of milk for cheese and yogurt, prior to drinking it as a beverage? Those both have less lactose per calorie. And for most of human history, those calories were something eagerly sought rather than avoided.

Evidently, many cats have lactase persistence.

Now, do you know of any nonmammalian animals with lactase?

Razib,

In the GWAS studies of height, did the genes for lactose persistence show up as having a large effect on adult height? (I suppose not since my recollection is that there were no genes of really large effect.)

It seems to me that lactose persistence gives people an easy way to continue to ingest large amounts of protein and calcium during their childhood and adolescence. I know a lot of, e.g., east Asian kids who never really like (cow's) milk, even when quite young, whereas other young kids seem to derive a huge fraction of their calories/protein/calcium from that one food item alone.

Do you know if there is any evidence that people figured out the use of milk for cheese and yogurt, prior to drinking it as a beverage?

i think so. people who have lower frequencies of LP have traditions of cheese and milk.

steve, interesting point. from what i recall the height studies are done on northern european population sets. including on finns, who have an non-trivial proportion of non-LP individuals, so i would assume it would show up. but perhaps the frequency in the sample was too low.

also, note that i think you can digest the fat and protein. it's just the carbs in the milk that go through. but as you indicate, the major issue might be the unpleasantness that the undigested carbs cause might dissuade kids from drinking a lot of milk.

There may be post-transcriptional or post-translational factors involved in lactase persistence vs. non-persistence. In an earlier study in which I was involved (Sebastio G, Villa M, Sartorio R, Guzzetta V, Poggi V, Auricchio S, Boll W, Mantei N, Semenza G. Control of lactase in human adult-type hypolactasia and in weaning rabbits and rats. Am J Hum Genet. 1989 Oct;45(4):489-97), lactase non-persistent subjects had 20 to 80% as much lactase mRNA as persistent subjects. Later studies, mainly by others, seemed to show a better correlation between enzyme and mRNA, so I had assumed something was maybe not quite right with our samples. But maybe we were in fact on the right track.
Also, if one takes tissue samples along the small intestine of rabbit or rat, the amount of lactase per amount of mRNA varies strongly (Keller P, Zwicker E, Mantei N, Semenza G. The levels of lactase and of sucrase-isomaltase along the rabbit small intestine are regulated both at the mRNA level and post-translationally. FEBS Lett. 1992 Nov 30;313(3):265-9). Partly this might just reflect amounts of intestinal proteases along the length of the intestine.
As an anecdote, the first chromosomal gene sequence for lactase was determined on clones of my own personal DNA. Sequencing a kilobase upstream of the mRNA start using my DNA (lactose tolerant) and then from a clone derived from a lactase non-persistent subject showed no differences that could explain the phenotype (Lloyd M, Mevissen G, Fischer M, Olsen W, Goodspeed D, Genini M, Boll W, Semenza G, Mantei N. Regulation of intestinal lactase in adult hypolactasia. J Clin Invest. 1992 Feb;89(2):524-9). In those days it would have been difficult to extend the sequence up to the SNP's discovered later (at about -13 kb).
Two comments on previous comments: I am not sure that cats are really all that lactose tolerant. In response to razib, drinking milk when lactose intolerant is more than just unpleasant. If you don't digest the lactose, bacteria in your large intestine will be happy to take care of it for you. The result is violent diarrhea and bloating.

In the case of my kids my son drinks tons of milk and is (roughly) > 99th percentile for height, whereas my daughter seems to like milk much less and is around 60th percentile. If I roughly estimate calcium and protein intake I can see huge differences, largely driven by milk consumption!

Maybe this has to do with acquired microbial fauna? I could imagine an interaction between genotypes that make it easier or harder for microbial gut communities to establish that affect lactase persistence.

I believe there are studies which show that Africans in particular, the ones who drink a lot of raw milk, have different SNPs for lactase retention than those found, and commonly tested for, in Europeans.

I am lactose intolerant, I don't have the right SNP version, I drink milk with no qualms or ill effects. Of course there are other lactase persistence SNPs out among Africans, some Europeans and SW Asians.

As for steve and his son and daughter. You can't extrapolate from two children and of different sex. How about switching to Soy Milk, non GM soy if you like, with added calcium. East Asians have been using it for millenia and look how many of them there are, and the Northern Asians are quite tall.

Cats drink cow's milk placed in their bowls. That does not make them lactose tolerant. Cat's cannot digest lactose, and too much milk makes them sick. Obviously someone does not take their cat to the vet and the cat is probably not neutered, and eating every small mammal and bird the cat can find depleting the environment of biodiversity.

Maybe having a little lactose over time stimulates the body to create lactase? I propose this hypothesis based on the alcohol model -- you stop drinking and your liver figures it doesn't need to produce the form of alcohol dehydrogenase that it used to when you were putting them back regularly (meaning a daily glass of wine with your daily dinner). The next drink you take after a very long dry spell certainly has a terrible effect; basically, if you want to keep drinking, you have to keep drinking, or your liver won't keep producing.

Re cats: Mine enjoyed lactose-free milk occasionally, when they decided water was not sufficiently interesting. It's silly to expect any adult mammals to be able to digest lactose; we humans are weird that some of us can. I can't, and my cats benefited from that.

Cats have shorter digestive tracts. Are they digesting the lactose in milk? Or just not much bothered by the failure to do so?

I guess the question about lactose-free-milk-drinking cats is are they digesting the galactose and glucose in that sort of milk? I have no idea. Maybe someone knows whether cat guts can absorb "*any* kind of sugars?

How Columbus was not a seer

razib — Wed, 13 Jan 2010 04:18:16 +0000

How Columbus was not a seer

A week ago I pointed out that in some visualizations of world wide population variation South Asians & mestizos seem to overlap which each other to a great extent. The reason for this is that both populations can be modeled as admixtures between two separate, but related, populations. Mestizos are the products of pairings between Europeans and indigenous America populations, while South Asians seem to be a stabilized hybrid population which emerged from the fusion of a West Eurasian (closely related to European) and East Eurasian (distantly related to East Asians) populations. The East Eurasian ancestors of South Asians may be distantly related to indigenous American populations, but in a world wide scale the relationship is relatively close (i.e., compared to Europeans vs. indigenous Americans). So when mapped onto a plot of genetic variation incorporating world wide populations South Asians and mestizos naturally resemble each other. That said, a commenter observes:

Great example of how two dimensions lose information.

Given how different the two populations are genetically, guarantee that the third component separates them pretty cleanly.

Correct. A new paper illustrates this. Magnitude of Stratification in Human Populations and Impacts on Genome Wide Association Studies:

Genome-wide association studies (GWAS) may be biased by population stratification (PS). We conducted empirical quantification of the magnitude of PS among human populations and its impact on GWAS. Liver tissues were collected from 979, 59 and 49 Caucasian Americans (CA), African Americans (AA) and Hispanic Americans (HA), respectively, and genotyped using Illumina650Y (Ilmn650Y) arrays. RNA was also isolated and hybridized to Agilent whole-genome gene expression arrays. We propose a new method (i.e., hgdp-eigen) for detecting PS by projecting genotype vectors for each sample to the eigenvector space defined by the Human Genetic Diversity Panel (HGDP). Further, we conducted GWAS to map expression quantitative trait loci (eQTL) for the ~40,000 liver gene expression traits monitored by the Agilent arrays. HGDP-eigen performed similarly to the conventional self-eigen methods in capturing PS. However, leveraging the HGDP offered a significant advantage in revealing the origins, directions and magnitude of PS. Adjusting for eigenvectors had minor impacts on eQTL detection rates in CA. In contrast, for AA and HA, adjustment dramatically reduced association findings. At an FDR = 10%, we identified 65 eQTLs in AA with the unadjusted analysis, but only 18 eQTLs after the eigenvector adjustment. Strikingly, 55 out of the 65 unadjusted AA eQTLs were validated in CA, indicating that the adjustment procedure significantly reduced GWAS power. A number of the 55 AA eQTLs validated in CA overlapped with published disease associated SNPs. For example, rs646776 and rs10903129 have previously been associated with lipid levels and coronary heart disease risk, however, the rs10903129 eQTL was missed in the eigenvector adjusted analysis.

The main point of the paper is to smoke out population substructure which might generate spurious false positives in health-related genome-wide association studies. The problem is pretty obvious. Imagine you have a medical study with a lot of blacks and whites, and you just assume they're all genetically basically the same. Then you look for associations of particular genetic variants within the population which has disease X. Of course, it could be that blacks or whites tend to have more of disease X than the other population, and, it turns out hat blacks and whites also tend to differ on a whole lot of genes. Modern human population genetics might have "disproved race," but it sure is very interested in "population substructure."

Patterns of between population variation can be visualized by extracting out the independent dimensions of variance, and plotting them against each other. Generally the charts I post on this illustrate the two dimensions which can explain the most variance in the data set (the alleles frequencies across all the SNPs in this case), principal components 1 and 2. But the comment above highlights that there are many other dimensions, though they explain less of the variance.

One issue that the authors of the above paper pinpoint is that the nature of these dimensions are sensitive to the populations which you include in your original data set to generate them. They distinguish here between the dimensions generated from the full HGDP data set, which includes ~50 world populations, and visualizations which rely only on one population. In this study they project their own samples of European, African and Hispanic Americans on the dimensions extracted out of the HGDP data set, and also onto dimensions generated from the populations themselves. As an example, consider Hispanic American projected upon the dimensions of variation constructed from Asians, Africans and Europeans, or, Hispanic Americans projected upon the dimensions of variation extracted from only the variance extant within their own population. From what I could tell they actually didn't find that correcting for total genome variation using these two methods was particularly helpful in generating greater clarity as to the role of population substructure in producing false positives. So let's focus on on the visualizations, which go back to the title of the post.

The first chart has PC 1 & PC 2 from the HGDP populations, with their sample of about 50 African Americans projected onto it:

Pretty much zero surprise here. I would be willing to assume that the self-identified African American who clusters with Europeans is an error of some sort (e.g., a sample mix-up), but other studies show the same tendency quite frequently. I conclude then that there are actually people who are inadvertently "passing" as black, at least culturally (on the outside they probably look whiter than G. K. Butterfield).

The second chart now has PC 1 & PC 3. So the dimension of variation which explains the second largest proportion of variance has now been replaced by the dimension which explains the third largest proportion.

Now Native Americans are distinct from East Asians in the HGDP sample. This is because of PC 3. This goes to the commenter's point that looking at more dimensions of variation gives us a better sense of real population differences.

Jumping back to PC 1 and 2, but with Hispanics projected onto the HGDP generated space:

I don't know the provenance of the Hispanics, but it looks to me that they're likely to include many Puerto Ricans, seeing as there's a large amount of African admixture here. Nevertheless, you still see the overlap between Hispanics and South Asians that you did with the Gujarati-Mexican comparison, though attenuated. So let's look at PC 1 & PC 3.

And yes, all of a sudden mestizos and South Asians do not overlap, and in fact South Asians are further from mestizos than Europeans or Middle Easterners. One could have predicted this from the previous chart.

Finally, I want to round out the inspection by looking at two charts which project European Americans onto PC 1, PC 2 and PC 3. The European Americans are black points.

Note that European American outliers seem to have a bias toward drifting in the direction of the Native Americans and African Americans. I don't discount the possibility of errors here, but it is important to note that deviations away from the HGDP European cluster in the last chart are toward the two groups which European Americans have historically been in contact with in North America.

Note: The subjects specific to this study seem to have been resident in the eastern half of the United States. This would tend to support my supposition that they are less likely to be Mexican Americans, and more likely to be Puerto Rican or Cuban Americans, if they were Hispanic.

Citation: Hao K, Chudin E, Greenawalt D, Schadt EE (2010) Magnitude of Stratification in Human Populations and Impacts on Genome Wide Association Studies. PLoS ONE 5(1): e8695. doi:10.1371/journal.pone.0008695

razib Tue, 01/12/2010 - 23:18

Tags

Population Substructure

I believe that in some contexts this is called "churning" -- the loss of information and approach to entropy. I've talked to linguists about whether something like this could happen to a language, making its antecedents unrecoverable. Two processes that happen to language are creolization and the development of a Sprachbund".

In the first, a language is stripped down to a minimum for contact with foreigners and made into a pidgin trade language which usually has vocabulary from two or several languages and a structure which is characteristic of most pidgins, but not necessarily of either donor language. The pidgin then becomes a creole when residents of the trade center grow up speaking mostly pidgin and the language develops beyond its rudimentary beginnings.

The Sprachbund is the trading back and forth of features between neighboring languages which are not historically related (from different language groups). One example is Romanian, which has picked up features from the surrounding Slavic languages -- the Balkan Sprckbund. Another is East Asia, where it is now thought that languages from several different unrelated or distantly related language groups (Austronesian, Tibeto-Burman, and maybe others) have picked up enough common features to make up a sort of adoptive family. The Sprachbund wikis are worth reading.

My theory was that of an unwritten language has endured, say, two cycles each of creolization and sprachbundization during (say) three thousand years (not impossible) ancestors before that time might be unrecoverable. It would essentially be a new isolate (as opposed to a survivor isolate).

The two processes are not unrelated, either. Every creole language would be part of a sprachbund comprised by its neighboring languages.

There's an example of this in the novel "The Good Soldier Schweik". Over the period of a century or more educated Czechs had picked up a German-type pronoun usage, whereas uneducated Czechs tended to stick to the Czech form, and nationalist Czechs insisted on the old form.

These factors might throw a monkey-wrench into attempts to build superfamilies larger than the known families (most famously Nostratic). They don't in any way discredit the established families (Indo-European, Semitic, Bantu, Malayo-Polynesian) but make work going beyond them difficult or impossible. The Turkish-Mongol-Manchu family has been questioned, though, and the language relationships of SE Asia are still up in the air.

Razib, I wonder whether it would ever be illuminating to plot (some) genetic data on a triangular diagram, rather than on multiple rectangular diagrams?
http://en.wikipedia.org/wiki/Ternary_plot

bio, follow the link to puerto ricans. there's a ternary plot in there.

Why don't people also do 3D plots? With computers, it really isn't that difficult.

And if the data isn't too dense, then taking a snapshot of the 3D plot from a small set of different angles should be quite telling. (Or, you could make interactive 3D plots, that could rotate, hide and show data, etc.)

they often don't look that good on 2-D paper i think. OTOH, seems like there'd be a good place for it in the supplemental information with visualization software.

Maple can make 3d scatter plots. You can move them around with the mouse to look at them from different angles.

Are Chinese subsets of Southeast Asians?

razib — Tue, 15 Dec 2009 06:23:35 +0000

Are Chinese subsets of Southeast Asians?

That's probably the big takeaway of a new paper on the genetics of Asians, a set which includes South Asians, but in the new research mostly focuses on the people of East Asia. In a global context this work is important. The backstory is that there are disagreements about the exact process of the "Out of Africa" migration. Most researchers would agree that the vast majority, perhaps all, of the distinctive genetic content of the human species derives from a migration from the African continent between 50 and 100 thousand years ago (closer to the former date than the latter likely). Note that there were other human lineages outside of Africa, the Neandertals being the most prominent, but various "archaic" groups were extant in eastern Asia as well down to the arrival of modern African-derived human groups. This is part of the reason why H. floresiensis isn't that outlandish, a lineage of H. er ectus was extant in Southeast Asia until the ~50,000 years ago, with the arrival of moderns.

Those are the agreements. The disagreement, in particular in regards to East Asia, is rather simple. Was there one, or two, waves from Africa, and did one, or both, settle East Asia? The two-wave model was promoted heavily in the early aughts by Spencer Wells. The whole argument is laid out in his book The Journey of Man: A Genetic Odyssey. The title hints to the fact that Wells and his collaborators primarily focused on paternal lineages, the Y chromosomes, in their reconstructions. Here's a screenshot from The Genographic Project which highlights the two-wave model:

In the context of East Asia, the two-wave model posits that there was a southern coastal migration, which pushed into Australia via southern India. And, there was a northern migration up through Central Asia from which arose both Europeans and East Asians.

The new paper in Science is Mapping Human Genetic Diversity in Asia:

Asia harbors substantial cultural and linguistic diversity, but the geographic structure of genetic variation across the continent remains enigmatic. Here we report a large-scale survey of autosomal variation from a broad geographic sample of Asian human populations. Our results show that genetic ancestry is strongly correlated with linguistic affiliations as well as geography. Most populations show relatedness within ethnic/linguistic groups, despite prevalent gene flow among populations. More than 90% of East Asian (EA) haplotypes could be found in either Southeast Asian (SEA) or Central-South Asian (CSA) populations and show clinal structure with haplotype diversity decreasing from south to north. Furthermore, 50% of EA haplotypes were found in SEA only and 5% were found in CSA only, indicating that SEA was a major geographic source of EA populations.

I don't think that the insight that language & genes are closely correlated will be that surprising. This was highlighted by L. L. Cavalli-Sforza on a coarser scale two decades ago, but has now been validated in a more fine-grained fashion by larger data sets and more powerful analytical techniques. The reasons for this should be obvious: marriage networks will coalesce around comprehensibility. There are exceptions, Tommy Lee's Greek-born mother reputedly did not speak English when she married his father, an American serviceman stationed in Europe.

To illustrate the relationships they constructed:

a) A maximum-likelihood based phylogenetic tree, showing relationships between groups

b) Also a Structure based chart which shows ancestral proportions inferred from 14 putative source populations

I've reedited and formatted figure 1 to fit on the screen clearly. The key here is to look how at the separation occurs as a function of language. This is not a magical process, no doubt language served as a barrier between groups in the past, and to some extent is also a cultural signal which can be used to infer the past identity of a given set of individuals in the present (this is an intuition which naturally less strongly perceived by Americans).

Note what I said in the title: the peoples of Northeast Asia are viewed in this study as subset a of various Southeast Asian groups. Additionally, the Indians branch out as the furthest outgroup, as we'd expect.

To get a sense of the relationships geometrically there is the obligatory PC plot. From figure 2 I want to focus on panels B & D. The first includes Indians and Europeans, while the second does not, and excludes the outliers among the East Asian populations. In other words, the first chart gives a sense of East Asian variation in a worldwide context, while the last is a finer-grained snapshot which elucidates the details of relationships among East Asians.

Remember that the axes represent independent dimensions which can be extracted out of the genetic variance data. PC 1 represents the dimension which has the largest component of variance, and PC 2 the second largest, and so forth. Much of nature of the scatter (or lack thereof) in this figure is predicable from previous work. Europeans are relatively homogeneous on a worldwide scale, but the Europeans represented by "CEU" are Mormons from Utah, who are themselves a subset of European variation. So the tight cluster is expected. Indians vary quite a bit along an axis. A recent paper has offered an explanation for why Indians so often seem to exhibit a linear distribution: South Asians can be conceived as a two-way admixture between a European-like population, likely invasive to the subcontinent, and an older resident population with distant affinities to the peoples to their east. This ancient eastern affiliated substrate, upon which the European-like element was overlain, is very evident in mtDNA lineages. "CN-UG" are Uyghurs, who are a relatively recent hybrid population.

The second panel is also somewhat expected. Recent work has reaffirmed a strong north-south cline within China among the Han. Chinese data have been confused, but these results seem to tacitly support the contention that the peoples of South China were often culturally assimilated and absorbed into the Han identity.

Finally, here's a figure which shows haplotype diversity declining as one moves north. A similar figure could be drawn along a line out of Africa to the northeast. Or from East Asia to North America. The inference is that the population with reduced diversity is derived from the population with greater diversity. The reason for this can be illustrated by the old example of a photocopy; subsequent copies have less information than previous copies. When a daughter population emerges from a parent population generally the former is a subset of the latter, and so it is less diverse. There may be a bottleneck whereby many distinctive alleles disappear through extinction. In this case the inference is that the populations of China, Japan, etc., are derived from a Southeast Asian group, ergo, they are less diverse. Not only that, but in terms of distinctive alleles the northern groups seem to be a subset of the southern groups. Here is the final paragraph of the paper:

To unambiguously infer population histories represents a considerable challenge...Although this study does not disprove a two-wave model of migration, the evidence from our autosomal data and the accompanying simulation studies...point toward a history that unites the Negrito and non-Negrito populations of Southeast and East Asia via a single primary wave of entry of humans into the continent

This conclusion rests on some assumptions. Here is the anthropology blogger Dienekes bringing up some objections:

As to the main thesis of the paper, namely that East Asians are descended from Southeast Asians rather than Central Asians, I have to say that I am not convinced. This thesis is based on two observations: minimum sharing between East Asians and Central/South Asians and south-north reduction of genetic diversity in East Eurasians. However, the high genetic diversity in Southeast Asians can be explained if they are taken to be old hybrids of Mongoloid northerners with "Australoid"-like southerners as physical anthropology suggests, and the seeming absence of influence of present-day Central/South Asians is due to the fact that the latter are largely Caucasoids of western Eurasian origin, and, thus, do not represent any putative ancestral populations to modern Mongoloids.

In other words, Dienekes is suggesting that the populations of Southeast Asia emerged in the same manner as those of South Asia, an admixture event between an indigenous substrate and an exogenous northern population. In China we have a great deal of historical evidence which points to north-south migration, in particular in the period between 500 and 1500. The greater diversity of the South Chinese may then derive in part from the fact that they are an admixed population, who carry within them the genetic heritage of the indigenous peoples, as well as Han immigrations from the north. But, it may also be true that the original Han were migrants from the south. One might posit the same with the more general model of East Asia, with the original Northeast Asians being derived from Southeast Asians, and contemporary Southeast Asians being admixtures between a "back-migration" from Northeast Asia and the local substrate. We know specifically that many of the peoples of Indochina have origins on South China, so this is not without some support from history.

Citation: Mapping Human Genetic Diversity in Asia, Science 326, 1541 (2009);
The HUGO Pan-Asian SNP Consortium, et al., DOI: 10.1126/science.1177074

razib Tue, 12/15/2009 - 01:23

Tags

Population Substructure

"back-migration" - a pleasant reminder of the Beaker People and their purported tendency to "reflux".

Heh. This might make the US Census category Asian as a natural grouping more justifiable.

Would it be likely that the ASI part of the Indian mix is closely related to the South-East Asian stock the East Asians evolved from?

sorry if this question shows gross ignorance, but what do the numbers represent in the phylogenetic tree?

hmmm and I thought Chinese were all descended from Indians! Or at least, thats what the Indian media seems to be reporting...

http://timesofindia.indiatimes.com/india/Ancestors-of-Chinese-came-from…

http://www.dnaindia.com/scitech/report_the-chinese-evolved-from-indians…

All speciousness aside, it seems there are any manner of (wrong) conclusions to be drawn from this data to suit any number of pet theories.

Would it be likely that the ASI part of the Indian mix is closely related to the South-East Asian stock the East Asians evolved from?

not closely related. let's go back to indians, who are ANI + ASI. if you have ANI, ASI, Europeans, East Asians, the ASI + East Asians would form one clade, the Europeans + ANI the other. the Europeans and ANI are rather close. the ASI and East Asians not so close. but closer to each other than they are to ANI & Europeans. or at least that's the model.

Stephen Oppenheimer has a single migration out of Africa

The green seems to be a Taiwanese aborigine (Austronesian) signature, and the only others who have ample amounts of it are Phillippines "negritos", Melanesians, and southern Chinese, which makes sense of as they would be admixed with the Austronesian phenomenon.

In microcosm, if this is to be interpreted by the HUGO team as Taiwanese aborigines as being derived from the more diverse Phillippines "negritos", then they mistook the buttock for the head.

sorry if this question shows gross ignorance, but what do the numbers represent in the phylogenetic tree?

pzed, those would be bootstrap values.

I'm still amazed at how "geographically correct" the PCA plots are. Europeans, Indians, East Asians fall almost exactly where they should be in a "real" map (however, Philippines / Malaysia are a bit off). This can be expected when we deal with close-by populations with constant, isotropic gene flow (e.g. within Europe). But between NW Europeans (CEU) and Japanese?...

Actually perhaps this is another sign that Europeans and East Asians really do originate from the same branch of Out-of-Africa migration. From this common origin, normal diffusion brings the familiar pattern of geographically correct PCs aligned with N-S and E-W axes.

Perhaps the fact that Malaysian/ Philippine Negritos do not fit the expected geographical pattern indicates that they do not come from the same branch, and therefore the PCA map (which is dominated by the other, EU-EA-India branch, since this has the largest variance) cannot be expected to apply to them?

</wild_uninformed_speculation>

thanks altyn!

My thinking is that east-Asians show strong adaptation for extreme cold weather, and that fits with the time of their split from the Caucasians (possibly due to a glacial intrusion in central Eurasia cutting off the two populations). Subsequent glacial maximum would have driven them south into SE Asia, overrunning the existing Australoid population there, and interbreeding to some extent. Then as the weather warmed they expanded north again. This explains why NE Asians are related to SE Asians but show less genetic diversity.

The timings would seem to be around ca 20-25,000 YBP for the Caucasian/east-Asian split, and around 10,000 YBP for the re-expansion northwards of the east-Asians.

Ashkenazi Jews are Middle Eastern & European hybrids

razib — Wed, 09 Dec 2009 16:20:19 +0000

Ashkenazi Jews are Middle Eastern & European hybrids

According to search engine traffic one of the most popular posts on this weblog has to do with the genetic background of Ashkenazi Jews. That is, those Jews whose ancestors derive from Central & Eastern Europe, and the overwhelming number of Jews in the United States. The genetic origins of this group are fraught with politics naturally. With the rise of biological science the characteristics of Jews were used as a way to differentiate them as a nation apart in more than a cultural and religious sense. After World War II other researchers attempted to show that Jews were not genetically distinct with relatively primitive blood group assays. Rather, they were the descendants of converts.

More recent genetic work has given mixed results. The reasonable inference then is that Jews themselves are a population with a complex history, and that complexity is manifest in their genetics. A new paper explores these issue in more detail, Genomic microsatellites identify shared Jewish ancestry intermediate between Middle Eastern and European populations:

Background
Genetic studies have often produced conflicting results on the question of whether distant Jewish populations in different geographic locations share greater genetic similarity to each other or instead, to nearby non-Jewish populations. We perform a genome-wide population-genetic study of Jewish populations, analyzing 678 autosomal microsatellite loci in 78 individuals from four Jewish groups together with similar data on 321 individuals from 12 non-Jewish Middle Eastern and European populations.

Results
We find that the Jewish populations show a high level of genetic similarity to each other, clustering together in several types of analysis of population structure. Further, Bayesian clustering, neighbor-joining trees, and multidimensional scaling place the Jewish populations as intermediate between the non-Jewish Middle Eastern and European populations.

Conclusion
These results support the view that the Jewish populations largely share a common Middle Eastern ancestry and that over their history they have undergone varying degrees of admixture with non-Jewish populations of European descent.

The general results of the paper are well illustrated by the figures.

The figure below shows the putative ancestry of individuals assuming a K number of ancestral populations. As you can see, the Jews within the sample are placed between Middle Eastern and European groups. At K = 5 and K = 6 the relationship between Jews and Palestinians shows up; a common ancestral population which parted aways at some point.

And here's a neighbor-joining tree. The Jewish groups in red, Europeans in blue and Middle Eastern groups in olive.

Now here are Jews compared to various populations. Jews are in red. I've reedited and labelled for clarity.

From the conclusion of the paper:

A simple explanation for the clustering of the Jewish populations is that this pattern is the consequence of shared ancestry with an ancestral Middle Eastern group. Under this scenario, the intermediate placement of the Jewish populations with respect to European and Middle Eastern populations would then result from early shared ancestry of the Jewish and Middle Eastern populations, followed by subsequent admixture of the Jewish populations that took place with European groups or other groups more similar to the Europeans than to the Middle Eastern populations in the study. Although it is difficult to assess the specific nature of the admixture on the basis of our analysis, this explanation is supported by other genetic studies that find a combination of shared ancestry and admixture among Jewish populations...and by historical records of conversions to Judaism...Further sampling of matched Jewish and neighboring non-Jewish populations will be informative for investigating the evidence for this scenario.

...

In several analyses, the population in the study that is most similar to the Jewish populations is the Palestinian population. This result is reflected by the fact that for K=5, Bayesian clustering with Structure assigns the Jewish populations and the Palestinians to the same cluster...and by the relatively close placement of the Palestinians and the Jewish populations in MDS plots of individual distances...This genetic similarity, which is supported by several previous studies...is compatible with a similar Middle Eastern origin of the Jewish populations and the Palestinians. Admixture of the Palestinians with groups with European origins might have maintained or augmented this shared ancestry, especially if it was paralleled with similar admixture of these groups with Jewish populations.

...

We note that caution is warranted in interpreting some of our results. For example, in the population trees produced from three distance measures...there is disagreement on the branching order of three of the European populations closest to the Jewish populations (Adygei, Sardinian, and Tuscan). Thus, from these data, it is difficult to make strong inferences regarding the most similar European populations to Jewish groups. However, consistent with studies that have incorporated a single Jewish population in a broader European context...southern groups from Europe are placed closer to the Jewish populations than more northerly groups.

This paper clarifies and puts into sharper focus what we knew, and leaves open more details for future research.

1) Jewish populations do have a common ancestral affinity.

2) But, that affinity is complemented by admixture with the populations amongst whom the Diaspora settled.

3) There is a suggestion that in the case of Ashkenazi Jews the European contribution was more likely to be from southern, and not northern, Europe. This is somewhat surprising in light of the fact that the Ashkenazi group crystallized during the medieval period in northern Europe, amongst German and Slavic speaking peoples. These data would imply that in fact there was a relatively strong separation between these groups and the Jews, at least when it came to gene flow into the Jewish group (other data from Poland does show the effect of Jewish assimilation into the gentile majority). Therefore, the admixture may have occurred within the bounds of the former Roman Empire, during the Imperial or early post-Imperial period.

4) The close relationship of Jews to Palestinians is not surprising. Jews are reputedly a Levantine population by origin, and the historical and genetic evidence points to Arabicization in the Levant and Mesopotamia as having occurred through acculturation, and not population replacement. Many of the Palestinians are likely of original Jewish or Samaritan origin, though I would guess that they were likely at least nominally Christianized during the Byzantine persecutions of the 6th century.

5) There remain questions as to which groups the Ashkenazi Jews admixed with, and when they admixed. There should be a different pattern of genetic variance if the admixture event was early and ceased, or if it was constant and gradual gene flow. The phylogenetics implies the former, because of the lack of much allele sharing with northern Europeans specifically, amongst whom the Ashkenazi Jews were resident for the past ~1,000 years. Within the text of the paper there are also hints of possible relationships to a population of the Caucasus, opening an avenue for some validity of the Khazar hypothesis. There have been other data which also point to the Khazar hypothesis. The origins of the Jews then likely are complex.

Many of the confusions and muddled points will likely be clarified soon with more data and analysis. At Dienekes' some observed that a greater number of Mediterranean populations would have been useful. What if the Jewish admixture event tended to occur with Greeks in Alexandria and in the cities of Asia Minor? That would explain the proximity to Italians, but lack of overlap with other European groups.

Citation: BMC Genetics 2009, 10:80 doi:10.1186/1471-2156-10-80

(H/T Dienekes)

razib Wed, 12/09/2009 - 11:20

Tags

Population Substructure

palestine before 1917
http://www.aldeilis.net/english/images/stories/israelpalestine/claimsbe…

I'm not surprised by this. I live across the street from two jewish families, one of them russian. there are alot of russian jews around here (Boston) and they look more russian than jewish.

these data don't show much slavic or germanic admixture. so if they look russian it is much more likely that they're simply half-jewish, as outmarriage rates were really high in russia during the communist period.

there are lots of russian jews here. some of them seem more jewish and some of them more russian. generally the more russian looking ones came later than the jewish looking ones, who tend to have come here earlier. my mother works at a jewish rest home and the residents kind of diss the russians for not knowing alot of the jewish customs, which were I guess discouraged under communism

wednesday I babysit and drink a little afterwards, sorry if above was a little disjointed

I have known a lot of Ashkenazi in my life, and the majority have blue or gray eyes and pale skin. That seems to suggest something other than a Middle Eastern and southern European mix. Maybe it is a founder effect.

selection? yes, the frequencies you see the literature suggest too much light eye color.

Is there a study like this which includes Sephardics and Mizrahis? It's odd that this study included so many middle eastern and mediterrenean groups but not those. Wouldn't that be crucial to answering the original question- "Genetic studies have often produced conflicting results on the question of whether distant Jewish populations in different geographic locations share greater genetic similarity to each other or instead, to nearby non-Jewish population."

Is there a study like this which includes Sephardics and Mizrahis?

this study includes 3 sephardic (or mostly sephardic) groups. i think there might be a study with persian jews coming out soon, so that would be mizrachi i guess.

One thing that keeps striking me in a lot of these studies is how little evidence of Khazar influence. Historically, the claim that the Ashkenazic population were descended mainly from the Khazars has been associated with anti-Semitism. In that regard, there's some degree of luck that the claim isn't accurate, since if it were I suspect many Jews rather than acknowledge the fact would have simply taken a very anti-science and anti-genetics attitude. (The tendency is very strongly for ideologues to accept science only when it doesn't contradict their pre-existing views). It should be interesting to see how in the long-run people react to this data. Many Palestinians don't want to be closely related to Jews. Moreover, Jewish supporters of Israel frequently like to argue that the Palestinians are a late population that came in during the Islamic conquest. So people on both sides of the Israeli-Palestinian conflict will be uncomfortable with these results.

Now personal speculation: If one looks at the late Second Temple period, around the birth of Jesus, it seems that the largest (or certainly one of the largest) groups of Jews were Hellenized Jews living in Palestine. It wouldn't be intrinsically surprising if they were part of the ancestral population of the modern day Palestinians. I'm not sure that your speculation about Samaritans makes sense. My impression is that they were always a small minority of the general population. There's been work on their genetics (see for example http://evolutsioon.ut.ee/publications/Shen2004.pdf ). I don't really know much about that sort of work but my impression is that the very small Samaritan population makes such work difficult.

re: khazar's, there's stuff about the adygei in there. i think it's probably not relevant, but who knows? if the palestinians have it might just be a common heritage of levantines and peoples of the caucasus. that being said, the original khazars were turkic, so that would show up pretty easily among jews.

yes. hellenized jews were probably a disproportionate number of the early christians. see rodney stark's work; the correlation between jewish settlements and early christian settlements is pretty strong. as i said, by the time the muslims conquered palestine it looks to have been mostly christian.

My impression is that they were always a small minority of the general population. There's been work on their genetics (see for example

way more numerous in the late antique period. analogy: zoroastrians. small population today, a lot more in the past. the samaritans were numerous enough to have rebelled during the reign of justinian the great and cause a lot of trouble. after that, not so much trouble :-)

re: sephardic jews. i should elaborate for future readers since semantic confusions tend to abound. after the expulsion of jews from spain they went everywhere. though many went to holland, or even to the ashkenazi jewish world of central europe, the largest number went to the muslim world (ok, the largest number stayed in spain and become christian). so the jews of morocco and the ottoman empire were quickly dominated by sephardic jews. some non-sephardic native jews persisted as minority traditions (romaniotes in greece), but basically the sephardic jews took everything over.

Razib,

Actually, if I'm not mistaken, in Morocco, the Sephardim formed a culturally elite minority that was numerically inferior to the pre-Sephardi Moroccan Jewry.

Actually, if I'm not mistaken, in Morocco, the Sephardim formed a culturally elite minority that was numerically inferior to the pre-Sephardi Moroccan Jewry.

sure. that's the problem with all the post-expulsion sephardic communities. but the quantities seem woolly to me.

There are more interpretations that are consistent with the historical record. I think there's an overreliance on the Levantine ancestry of these groups.

I recall reading that approx. 10% of the Roman Empire was "Judaic" -- about half of the Roman Empire were Phoenician/Semitic areas, from Spain across the Mediterranean to Lebanon. We know that these areas were also later on heavily Jewish -- and one can reasonably expect that at least proto-Judaism and allied religious traditions went along the trade routes through these regions.

So maybe these "admixtures" aren't late admixtures at all -- but reflect ancient population links on the Mediterranean coasts and islands, that became specifically "Jewish" later, as the lines between the Hellenic Christians and non-Hellenic Jews left little LEGAL space for all the other varieties of Hellenic/Semitic religious expression.

Both Jews and Christians have religious reasons to forget the complexity of pre-Talmudic/pre-Nicene religious expression. But the data here should remind us of it.

One must remember that apparently the Urheimat of Afroasiatic languages is either Ethiopia or the current Sahara desert -- that there were millenia of trans-Mediterranean trade of the culturally and linguistically related Egyptian populations -- that the Jewish populations were part of the greater Phoenician trade networks that lasted until the Carthaginian war -- that the seminal "Jewish" identity as opposed to Hellenistic is a late event, remember the Maccabean forced circumcisions -- that even within the Torah tradition, the Samaritans were a large population until the 10th century persecutions in the ME -- that Rome itself was a babble of cultures, fusing Semitic traditions with Hellenistic traditions willy nilly for close to a millenium -- that ambiguous, marginal cultural populations exist to this day, including such groups as the Mandaneans.

Those trees seem to reflect that reality -- that what we see is populations crystallized out of the Mediterranean stew; particularly note the connection between Anatolians ("Turks") and Levantine-related populations. I think we're being fooled by people's self-image.

Orcadians is an odd choice for germanics. Why not germans? There's a good chance ashkenazim are closer to southern europeans on these graphs simply because there's a cline from northern europe down the middle east, I don't think this establishes southern european blood in them at all.

Orcadians is an odd choice for germanics. Why not germans? There's

there are germans in the HGDP sample?

Joshua, among Jews who will matter in the next generation (Orthodox Jews and the right-wing fringes of others) Khazar findings are far less emotionally repugnant than Palestinian findings. Khazars, after all, were spoken of highly by Judah HaLevi, were righteous converts (or so the right-wing masses believe) and were Jewish warriors, a thing the grandchildren of Holocaust survivors are highly supportive of. Palestinians on the other hand...

Being one of those people (on the fringes of non-orthodoxy and considering Orthopraxy) I want to quickly note that in every sense that matters on the emotional front of this matter, Druze are not Palestinians. In fact, I'm hesitant to pass around this link to friends on account of the fact that no non-specialist will understand that Razib seems to have included Druze and Beduin within the bubble of "Palestinian" being as he's more concerned with geography than with the squables among various micro-peoples on some tiny patch of land.

Knowing their confusion on this score I may simply attempt a quick turnabout by pointing them towards my Lechi-fighter friend Ezra Yakhin who wrote a whole book on how the Palestinians are the literal descendants of Amalek (and why they should therefore be ritually slaughtered). Horror and laughter aside (laughter because it's the outlandish suggestion of an old guy who I saw shouted off the stage for this suggestion by Lechi colleagues of his at a program in honor of Lechi's founder, Yair Avraham Stern) this point will in fact alleviate loads of psychic discomfort by reminding post-Holocaust Jewry of the ancient Jewish approach where relativity wasn't everything and the descendants of Esau were considered to be the most despicable anti-semites in the world.

On to subjects on which I'm a moron: Razib, any chance you could explain briefly how to understand all of these pretty pictures?

Seriously, I know I can read a few wiki entries on em and I already have some small grasp of a few of them but from long experience (not least by your almost wholly accurate descriptions and explanations of matters in my own field of Jewish arcana, where your level-headedness demonstrates how easilly accessible accurate analyses is of ANY issue so long as your love of exotica is married to skepticism) I've come to trust you for accurate and pithy descriptions of things. Any chance you could give me a few sentences offering the basics on each of these different types of graphs. It would actually make quite the awesome post that you could perpetually link to when including graphic material in any analysis. I'm certainly not the only intelligent auto-didact who would be able to more greatly appreciate your pieces were he to have more than a general guess as to how to appreciate the data in your colorful graphs.

Cheers,

mnuez

Blogged on a phone

There's plenty more I'd like to talk about with regards this post such as which population of Palestinians were chosen (Hebron or Khan Yunis?) And whether the chosen population was selected for based on their hypothesized antiquity in the area (I know black-African and Egyptian "Palestinians" living in Jerusalem, were they as likely to have been picked for study as a Gallillean Palestinian with a 1200 year old family tree?) And of course many other related topicas as well but I believe that even at this juncture I can safely claim my award for longest comment left by a phone tapper.

I want to quickly note that in every sense that matters on the emotional front of this matter, Druze are not Palestinians. In fact, I'm hesitant to pass around this link to friends on account of the fact that no non-specialist will understand that Razib seems to have included Druze and Beduin within the bubble of "Palestinian"

figure D = just palestinians, so i don't know where you'd getting that i confounded the two. i specifically didn't want to talk about the druze because they're a relatively genetically isolated group like the kalash who seem to have 'drifted' into their own direction. and there's a huge difference between syrian, lebanese and israeli druze on the questions that someone like you would care about :-) (i.e., "but is it good for the jews?") anyway, i'd be willing to bet $100 dollars that modern palestinians share more distinctive alleles with jews circa hasmonean period than the ashkenazi. in other words, modest confidence, though i wouldn't be surprised if i was wrong either. perhaps the gov. of israel will fund some DNA extraction? thank god jews didn't cremate!

as for the plots, my understanding is that they're just plotting genetic distances between individuals based on the distinctive alleles shared. you could represent it 3-d too i guess. basically you have a matrix where the rows & columns are ij pairwise distances between individual i and j and you're visualizing into a particular space of independent dimensions.

p.s. by chance i saw you making fun of "fred scrooby" via a google search. pretty funny stuff. interesting how two jewish grandparents becomes 'some distant jewish ancestry.'

re: MDS plots, they used the cmdscale function in R:

http://www.personality-project.org/r/mds.html

The results of this study are consistent with the findings in my book "The Jews of Khazaria, Second Edition" - the only comprehensive book on the Khazars that incorporates genetic data. If you were annoyed by the pseudohistorical book denying Ashkenazim have any substantial Israelite ancestry that recently got undeserved press in the New York Times and Wall Street Journal, you can refer to my book instead. There is nothing offensive about the fact that Southern Europeans and Khazars (and others) have contributed to Ashkenazic ancestry as well.

Is the Jewish affinity with Adygei truly Khazarian or is it because of Armenian/Anatolian-Turk DNA the Ashkenazi and European Sephardim picked up in Asia Minor?

Bauchet in 2007 demonstrated the Ashkenazim clustered quite closely with Greeks and Armenians.

"there are germans in the HGDP sample?"

That's the problem, there aren't. I see a lot of comments here saying that this proves there is very little germanic and slavic blood in ashkenazis, but where does that come from? It's obvious with slavics, because we have russians represented. But how can it be claimed that there's little germanic blood in them?

joe, most euro variance is north-south. in the context of the questions here orcadians are probably a sufficient proxy.

That's true, but I'm still not impressed. Doesn't that cline keep going right down the edge of the Mediterranean? Shouldn't a jewish-north europe mix looks an awful lot like a jewish-south europe mix but with slightly different weights?

The Khazars were not exactly an East European or a slavic group. They were a people hailing from Northern Caucasia with at least some Central Asian blood. Ashkenazis may be genetically close to Kurds, Turks and Armenians, however Chechens have also been proven to be genetically close to other Caucasian peoples including Armenians. Modern day Chechnya would have been part of Khazaria. It also would include Ingushetia, Ossetia, Dagestan, etc. Middle eastern genetic markers have been found in those groups. The Ashkenazis also have assorted Haplogroup G markers, which are found near or around the Caucasus, some originating in the Caucasus perhaps. Although, in depth studies have not been done in the caucasus region, especially in the northern cacausus. So the fates of the infamous Khazars still remain unclear.

Joshua said:
One thing that keeps striking me in a lot of these studies is how little evidence of Khazar influence. Historically, the claim that the Ashkenazic population were descended mainly from the Khazars has been associated with anti-Semitism. In that regard, there's some degree of luck that the claim isn't accurate...

But there are studies out there which detail exactly the genetic contribution of Khazars to Jews?! Check out this study by Ellen Levy Coffman, who pretty much prooves that Ashkenazi Levites are Khazars by descent:
http://www.jogg.info/11/coffman.htm

The genetic similarity of Jews to Kurds and other populations of the northern Middle East was posited, if I recall, in a Y-chromosome study by Nebel. Kurdish and Anatolian samples have not so far as I know been included in any of the more recent Jewish autosomal studies, and Armenian samples have only shown up in Bauchet's paper from 2007. So it's really not clear how close Jewish populations are to the Kurds or Anatolians at this point, and we should be wary to posit a relationship since the autosomal studies give a picture of significantly more admixture (at least among Ashkenazim) than the Y-chromosome studies.

I have nothing against the Khazars (in fact, I considered the possibility of being their descendant rather intriguing as a child until I learned of their propaganda utility to Stormfronters and Arab nationalist crazies). However, it would be nice to actually know a bit about their genetic makeup before considering any relationship between them and the Ashkenazim to be proven. At its strongest, the case is very circumstantial.

once the sample sizes get huge for all populations does the signature of rare "east asian" haplotypes in ashkenazi jews rise above that of sephard, mizrachi and other west european populations? if so, that might be a signature of turkic admixture (the khazar themselves were an admixed horde, but their origins were certainly in greater mongolia, like the avars).

The historical evidence in favor of Ashkenazi and European Sephardic DNA being mostly Anatolian and not Levantine looks quite good to me.

Historically we know Jews enjoyed a large presence in Asia Minor reaching all the way back to the 4th century B.C. and lasting to the post Western Roman Empire-Byzantine period.

Since the Jews of Asia Minor were established there for more or less 1,000 years I don't see how it is possible that the ancestors of European Jewry could have avoided picking up a considerable amount of Anatolian-Turk, Greek and Armenian DNA from:

A) Jewish men taking local Anatolian wives and

B) from Anatolian converts to Judaism when the Jews actively engaged in proselytism before Christian authorities outlawed Jewish proselystism between the 5th and 7th centuries.

There is no doubt some Levantine Arab blood left in modern European Jewry. However I would bet that the majority of their Middle Eastern DNA is Anatolian-Turk, Armenian and maybe Kurdish. Probably only a moderate to low amount of European Jewish DNA originates from the original Levantine Jews.

Even the above paper states that Turkish Jews were a bit difficult to distinguish from Ashkenazi Jews, and it is certain Turish Jews have a substantial Anatolian DNA component.

no turk. substitute in kurd (or as they would say, 'isaurian').

observer's comments seem plausible enough to me. but one should note that on the Y chromosomal lineages there is a clear connection to *levantine* middle easterners, not anatolians. i.e., the cohen modal haplotype is from the branch of haplogroup J common among arabs. though the kurds are collapsed into this group too, though the turks are not. see more at the citations in wiki:

http://en.wikipedia.org/wiki/Jew#Male_lineages:_Y_chromosomal_DNA

though this is consistent with the model above, as long as patrilineages are maintained, the autosome can be replaced....

The Khazar theory may seem improbable to certain political groups and ideologues, however let us not mix politics and genetics. There needs to be larger studies done in the greater Caucasus region, there is a lot of data that is missing for the area. We should compare how the Chechens, Dagestanis, Ingush and other Caucasian peoples compare to Ashkenazis first before making any rash conclusions either way.

me. but one should note that on the Y chromosomal lineages there is a clear connection to *levantine* middle easterners, not anatolians. i.e., the cohen modal haplotype is from the branch of haplogroup J common among arabs.

This is a good point.

But J2 - which is common in Ashkenazi and European Sephardi - is found most frequently in the Caucasus (for "mike" and his Khazars), the Fertile Crescent and Anatolia and gradually less as one moves Southeast from Syria and Lebanon and deeper into Arabia:

http://en.wikipedia.org/wiki/Haplogroup_J2_(Y-DNA)

Haplogroup J2 is found mainly in the Fertile Crescent, the Caucasus,[5] Anatolia, the Balkans, Italy, the Mediterranean littoral, the Iranian plateau, and Central Asia.[1] More specifically it is found in Iraq,[20] Syria, Lebanon,[21] Turkey, Israel, Palestine, Greece, Italy and the eastern coasts of the Iberian Peninsula[22], and more frequently in Iraqis 29.7% (Sanchez et al. 2005), Lebanese 25% (Semino et al 2004), Palestinians 16.8% (Semino et al 2004) [1], Syrians 22.5% (Luis et al. 2004), Sephardic Jews 29%, Kurds 28.4%, Jordan 14.3%, Oman 15% (Di Giacomo et al. 2004) & 10% (Luis et al. 2004), UAE 10.4%, Yemen 9.7%[23], in Israel[1], in Palestine[1], and in Turkey.[4]

J2 is found at very high frequencies in the peoples of the Caucasus - among the Georgians 21%[5]-72%,[6] Azeris 24%[7]-48%,[6] Ingush 32%,[8] Chechens 26%,[8] Balkars 24%,[13] Ossetians 24%,[8] Armenians 21.3%[6]-24%,[8] and other groups.[5][8]

In Europe, the frequency of Haplogroup J2 drops dramatically as one moves northward away from the Mediterranean. In Italy, J2 is found with regional frequencies ranging between 9% and 36%.[14] In Greece, it is found with regional frequencies ranging between 11% and 46%. Frequencies are high in Turkey, approximately 24% of Turkish men are J2 according to a recent study,[4] with regional frequencies ranging between 13% and 40%.[12] Combined with J1, up to half of the Turkish population belongs to Haplogroup J.

It has been proposed that haplogroup subclade J2a-M410 was linked to populations on ancient Crete by examining the relationship between Anatolian, Cretan, and Greek populations from around early Neolithic sites[24]. Haplogroup J2b-M12 was associated with Neolithic Greece (ca. 8500 - 4300 BCE) and was reported to be found in modern Crete (3.1%) and mainland Greece (Macedonia 7.0%, Thessaly 8.8%, Argolis 1.8%) [25].

Sephardic Jews have about 29% of haplogroup J2[1] and Ashkenazi Jews have 23%[1], or 19%[26]. It was reported in an early study which tested only four STR markers [27] that a small sample of Italian Cohens belonged to Network 1.2, an early designation for the overall clade now known as J2a4, defined by the deletion at DYS413. However, a large number of all Jewish Cohens in the world belong to haplogroup J1 (see Cohen modal haplotype).

J2 subclades are also found in Iran, Central Asia, and South Asia.

Haplogroup J2 has been shown to have a more northern distribution in the Middle East, although it exists in significant amounts in the southern middle-east regions, a lesser amount of it was found when compared to its brother haplogroup, J1, which has a high frequency southerly distribution. This suggests that, if the occurrence of Haplogroup J among modern populations of Europe, Central Asia, and South Asia does reflect Neolithic demic diffusion from the Middle East, the source population is more likely to have originated from Anatolia, the Levant or northern Mesopotamia than from regions further south.

For more on the Cohen Modal Haplotype (CMH), check this out:
http://dienekes.blogspot.com/2009/08/finally-updated-look-at-y-chromoso…

Basically it seem that the original Cohen only lived in Roman times at the earliest - per Dienekes.

Nebbish said:
I have nothing against the Khazars (in fact, I considered the possibility of being their descendant rather intriguing as a child until I learned of their propaganda utility to Stormfronters and Arab nationalist crazies). However, it would be nice to actually know a bit about their genetic makeup before considering any relationship between them and the Ashkenazim to be proven.

So to simplify, you are in favor of any science that supports certain Jewish causes and against it if it doesn't - basically you view science as a tool of politicians - nice?!

At its strongest, the case is very circumstantial.

"There are none so blind as those who don't want to see..."

The fact that the Khazar elite converted en masse to Judaism is not in dispute, the only fact in dispute is if the rank and file Khazar converted too. Of course it would be nonsense to say that Ashkenazim are simply Khazars, they are not, but they have substantial (12%) Khazar admixture.

Additionally, I share genomes with a number of Ashkenazim, and can tell you that the a many of them have minor East Asian gene segments - usually 1-2% East Asian.

The 12 or so percent refers to Eastern European admixture. The Khazars were not Eastern Europeans. They were from the Caucasus. Caucasians have been proven to have y-chromosomal lineages similar to Middle Easterners. Ashkenazis (a caucasian people) cluster closely with Armenians in fact.

*I should have written Armenians (a Caucasian people) cluster closely with Ashkenazis in fact.

careful about the terms thrown around guys. remember that for most of history (up until the early 20th century genocide) armenians were more of an east anatolian populations than a caucasian one. as for the khazar's being caucasian, their locus of power was on the broad plains to the north, and though they were likely an admixed group by the time they began to be influenced by judaism their origins are turkic, and so ultimately in mongolia.

pconroy,
That isn't what I meant at all. The maximalist position that would support the Jewish ethno-nationalist political cause is no admixture of any kind. Not even the original Hammer paper from 2000 proposed that, although it and the other early Y-chromosome studies claimed that the situation was close to that. The autosomal studies clearly show that admixture has occurred, and I am quite comfortable with that. I am not comfortable with declaring all or even a substantial portion of the admixture to be of Khazar origin given that we know so little about the genetics of the Khazars (almost no physical remains, lack of clarity about their population genetic profile at the time of their conversion to Judaism, etc.). It seems really premature to declare the case of Ashkenazi admixture solved based on the circumstantial evidence that Ellen Levy Coffman cites. There are potential alternative explanations for small quanta of East Asian ancestry among Jews (silk road trade, a deep ancestral component within the Middle East, etc.), although the Khazars might be the answer. And there is a nontrivial chance that extensive admixture with the Khazars (whose origins lay in Mongolia, as Razib has noted) would not explain where Ashkenazi and other Jewish populations lie in relation to S. Euro populations.

As for the relationship between Turkish Jews and Ashkenazim, it doesn't necessarily say much about Jewish Anatolian admixture. The Turkish Jewish population is heavily Sephardic in ancestry due to migrations after the expulsions from Spain There may not be much continuity with the ancient Anatolian Jewish populations. Having said that, Roman-period Anatolia, Greece, and Italy rank high on my likelihood meter for sources of admixture that might differentiate modern Jews from the peoples who currently inhabit their ancient stomping grounds due to large Jewish populations in those regions from the Hellenistic period onward.

The Khazar theory may seem improbable to certain political groups and ideologues, however let us not mix politics and genetics.

It is hard for people to separate them in this context. Frankly, I find claims of legitimacy based on ancient ancestry to be very hard to see as philosophically justifiable anyways so it isn't that relevant. However, it isn't unreasonable to look at the data and wonder what political implications it has. The key there is to be careful that the genetics informs ideology or politics not the other way around.

I believe that the Khazar theory may actually have some merit. If you were to look at the Anatolian example, you would find that the majority of Anatolian "Turks" are actually Turkified Anatolians. The same could easily have happened for the Khazars. The majority of "ethnic" rather than racial Khazars could have been native Northern Caucasia, and the central asian admixture would have been at a minimum. Western Asian haplogroups are very common in the Caucasus region. Just as historically many Anatolians have considered themselves turks until population genetic studies proved otherwise. Both J1 and J2 appear in the Caucasus, as do various Haplogroup G subclades which Ashkenazi jews belong to, although this still needs more testing. The big mystery so far deals with haplogroup E1b1b1. It has not been tested thoroughly in the Caucasus, especially the Northern Caucasus. The Khazar question is still an open one.

There must be something about the Jews to be so intensely studied. I think it is not so much genetics but the fact that they have kept fairly distinct through the millenia. What other ethnic group has done so? No more Romans, we can only guess at the Hellenes of old, Etruscans gone, Phoenicians disappeared, even the Franks, Angles, Saxons, Jutes, Celtiberians, Vandals..the list goes on. All gone, but the Jews are still with us. That says something about them and probably why they are so disliked by just about everyone.

I don't really accept this separation thing between Europe and the Middle East in genetics. The average European and average Middle Easterner just a cline in the West Eurasian Caucasoid form of humanity. Most Europeans derive from people with Middle Eastern origins whether it was in the Neolithic or later ages.

With SNP tests you can separate ethnic groups, Northern English from Southern English, people from Watford from people from Westminster. The answer is so what, a big deal. It depends on what choices you make, which SNPs vary more in each population. Jews would overwhelmingly cluster with standard Europeans compared with South Asians or East Asians or Africans. Jews would cluster with Italians, Greeks and other Mediterraneans compared with NW or NE Europeans. Jews will cluster with Lebanese and Syrians if they were the only ones chosen. It would be interesting to see where Greeks, Italians and other Mediterraneans cluster with when only Middle Easterners are chosen. Choice rules which cluster they fit into.

Interesting to see that the researchers have mostly Jewish names. Probably means Jews want to claim their distinctness and their origins to the Hebrews, Israelites, Judaeans and other mythic ethnic groups to stake their claim to Israel/Palestine. Lots of agendas there.

Here is Dienekes' take:

http://dienekes.blogspot.com/2009/12/jews-intermediate-between-middle.h…

Dienekes said...
The more interesting thing (to me) about recent Jewish demography is how Ashkenazi Jews (who were a very small minority of the Jewish population) grew to become a great demographic component of that population.

How did this growth occur? the AJ found new living space in Europe, where they could grow in numbers. By contrast, Jews elsewhere had already experienced their growth phase and faced decline due to conversion. Thus the AJ component of the total Jewish group increased.

Drawing a parallel, Jews in Palestine in the Hellenistic-Roman era were plentiful (locally) but they were demographically constrained by the limited available space. But, their Diaspora had much more available space to grow in. So, I would guess that the major part of extant Jews' ancestry is from Roman-era Diaspora Jews rather than Roman-era Palestine Jews.

The interesting question (to me) is how much European ancestry did the Jews receive in the Hellenistic-Roman period, and how much in the medieval-Roman one. I believe that both elements are present, but I would wager that the European element in AJs consists of a fairly uniform Italian-Balkan-Anatolian "old" stratum, and a very variable German-Slavic "new" stratum.

Thursday, December 17, 2009 11:43:00 AM

This is what can happen when geneticists have their ideologies determine where to look and what to look at specifically, while ignoring other possibilities. I like how the Khazars are referred to mainly as R-M17 by the researchers, when in the Northern Caucasus region multiple Western Asian haplogroups are clearly present. The Khazars would not have been merely Eastern Europeans. Just take a look at the Caucasian peoples now living around the area of what was Khazaria. There are striking similarities with the Ashkenazi Jewish population.

Here are my thoughts on the the origins of the Ashkenazi, in light of the genetic data presented so far.

1. The "Near eastern" populations utelized are not representative specimens. They are levant populations, which, as historical records demonstrate, received substantial European admixture over a 1300 year period (at least). They also admixed with Turkish and Armenian populations. Cluster analysis of the Levant populations demonstrates this in a number of studies. According to Cavalli-Sforza (2008), the Druze and Palestinians are around 50% European, and 30-50% Near Eastern (my estimates from the structure chart), with some admixture from Central Asian Caucasoids.

The results of genetic affinity studies using Palestinians and Druze as proxies for Near Eastern populations, are therefore different from what researchers have reported. Rather than showing how "Near Eastern" the Ashkenazi are, these studies demonstrate how European the levant populations are. Notice that in the occasional studies which do include surrounding Near Eastern populations, the Levant clusters closely with Europeans. This is exactly what we should expect from the historical record - hybridized (Europeanized) populations with a Near Eastern base.

2. The similarity of Ashkenazi and Sephardic Jewish populations in the above study, is misleading for two reasons. (1) The Sephardic populations are European in origin, and do not represent Near Eastern Jews. (2) The Ashkenazi population is likely drawn from the same base as the Sephardic population of Spain. The implication, then, is that Sephardic and Ashkenazi populations are not totaly distinct, but instead, variants of a common population with origins in South Eastern Europe, where Jews first appeared on the continent. Actual Near Eastern Jewish populations from Iraq, Iran, and other countries, do not cluster well with Ashkenazi Jews. Had they been included, the results of the above study would likely have looked much different.

3. Ashkenazi Jews fit well with South Eastern Europeans, particularly Greeks and Italians. The reason why is obvious: according to the historical record, Greek and Roman converts to Judaism WERE the source of Jewish introduction to the continent. Furthermore, according to historical evidence, there was not, at any time, a major movement of Palestinian Jews into Europe. Now, the question is, does the genetic evidence support the historical documentation? Yes, it is almost exactly what you would expect to find. Here is why:

A. Historicaly, Ashkenazi were founded from Roman and Greek converts, so their affinity to Europeans should be closest to those populations. The genetic distance should be low. It is. In the above study, Ashkenazi only segregate from other Europeans at clusters of k=5 or more. But notice that they NEVER group well with Palestinians (only 25-30% it seems), and that this grouping dissappears altogether in the next iteration. The temporary grouping is explained by the European admixture in Palestinians. Other studies show similar structuring of Ashkenazi clustering with Greeks and Romans.

B. Ashkenazi have some influence from Kazars according to historical records, but not much. It's also likely they have some admixture from Caucaus region peoples and surrounding populations such as Turks. This explains the weak affinity to the Adgei.

C. The mysterious "third component" of the Ashkenazi is probably an artifact of several genetic factors other than admixture. Notice that the "mysterious" Jewish signature (at k=5+) isn't found in Levant populations, despite both Jewish and levant groups having substantial admixture with Europeans and partial origin in the SAME Near Easterners. The mysterious signature has been found in most other genomic studies of the Ashkenazi. Notice that the "Jewish" component is highest in the Sephardic populations. Why? Because the Jewish component is probably ancient genetic variation from South Eastern Europe that was brought by Sephardics to Spain, but substantialy lost by the Ashkenazi through (1) drift, (2) bottle necking, and (3) admixture with non-Southern Europeans. This is supported by the fact that the Southern Europeans (Italians, etc) have the highest values for the third component than all other European populations. Only the Adgei have higher a greater proportion, and they are in the caucaus. This seems odd at first. The explanation may be that the Caucaus, Northern Turkey, and Southern Europe, share a common genetic affinity which is reflected in the structure chart. This affinity may exist on a cline so that the southern-most and eastern-most portions of Europe share it more than the rest. If Jews were drawn from this region, then it would explain the observed population structures. Of course, there may be some other explanation too, this is rather speculative.

Your thoughts?

To be Coloured in South Africa means being all of the above

razib — Sat, 05 Dec 2009 19:53:29 +0000

To be Coloured in South Africa means being all of the above

About six months ago I had a post up on the Cape Coloureds of South Africa. As a reminder, the Cape Coloureds are a mixed-race population who are the plural majority in the southwestern Cape region of South Africa. Like the white Boers they are a mostly Afrikaans speaking population who are adherents of Reformed Christianity. After the collapse of white racial supremacy many white Afrikaners have argued that it is natural and logical to form a cultural alliance with the Cape Coloureds because of the affinity of language and faith (Afrikaans speaking Coloureds outnumber Afrikaans speaking whites). For their own part, though a people of color who suffered under Apartheid, the Coloureds have an ambivalent relationship with the black majority and have supported several white dominated political movements since the end of Apartheid.

One of the reasons that the Afrikaners and Coloureds have a relationship is a genetic one: the white ancestry of the Coloureds is the same melange of Dutch, French Huguenot and German which gave rise to the Afrikaners. Though I am unaware of modern genetic studies, older genealogical research has concluded that on the order of 5% of Afrikaner ancestry is non-white, and almost certainly it is through the "passing" of Coloureds into the Afrikaner population. As the Coloureds share language and religion with the Afrikaners this would naturally not be particularly difficult if they could pass themselves off as unmixed European. Or, at least if on the frontier farms of the 18th and 19th century if the neighbors did not inquire too closely as to the provenance of an individual whose ancestry was probably mixed. It is because of the non-white genetic load, small, but not trivial, that individuals such as Sandra Laing emerged from Afrikaner pairings during the era of Apartheid.

But though the Cape Coloured derivation from Europeans and local Africans, a mixture of Bushmen, Khokhoi and Xhosa, is well known and attested, there are other groups who are in the mix, quite literally. The Cape Colony was a way station between the far flung holdings of the Dutch East Indies Company, the VOC, one of the first major joint-stock corporations in the world, and the mother country. Just as Europeans arrived from the north, so Asians were brought from the east, though in this case they were generally slaves or bonded laborers. Large numbers of Southeast and South Asians arrived to serve the Dutch and provide labor which the locals would (because they were free and outside of colonial control) or could not (they died of disease and maltreatment). The Cape Malay community, which is in some ways affinal to that of Cape Coloureds, serves as a cultural testament to the Muslims amongst those brought to Africa. But the Cape Coloureds themselves no doubt have Asian ancestry which has left fewer salient cultural marks than that of their European ancestors.

But recent genetic data is clarifying that ancestry. It seems plausible to assume now that Asian ancestry surpasses European ancestry among the Cape Coloureds, with African ancestry still retaining a plural majority. A new paper in Human Molecular Genetics confirms earlier findings, Genetic structure of a unique admixed population: implications for medical research:

Understanding human genetic structure has fundamental implications for understanding the evolution and impact of human diseases. In this study, we describe the complex genetic substructure of a unique and recently admixed population arising 350 years ago as a direct result of European settlement in South Africa. Analysis was performed using over 900 000 genome-wide single nucleotide polymorphisms in 20 unrelated ancestry-informative marker selected Coloured individuals and made comparisons with historically predicted founder populations. We show that there is substantial genetic contribution from at least four distinct population groups: Europeans, South Asians, Indonesians and a population genetically close to the isiXhosa sub-Saharan Bantu. This is in good accord with the historical record. We briefly examine the implications of determining the genetic diversity of this population, not only for furthering understanding of human evolution out of Africa, but also for genome-wide association studies using admixture mapping. In conclusion, we define the genetic structure of a uniquely admixed population that holds great potential to advance genetic-based medical research.

The name of the game again is population substructure in the service of medical research. In this case, the Cape Coloureds are relevant because of their highly diverse origins, and so their possible relevance for admixture mapping. If a genetic variant which causes disease differs greatly in frequency between two distinct populations, an admixed population would be ideal to explore that particular relationship. African Americans are a case in point, they are modeled easily as a two-way mixture between Northern Europeans and West Africans. The Cape Coloureds present a more complex case, because their admixture derives from several different very distinct groups.

Historically the Asians seem likely mostly to have been from the Indonesian archipelago, as well as from coastal southern India and Bengal. The African demographics of the Cape have changed considerably since the colonial era. The Khoisan groups, both Bushmen and Khoikhoi, have declined in numbers in relation to the Xhosa. Though the Xhosa themselves have clearly absorbed a great deal of the Khoisan, before, and after, that period. For this study the researchers used 20 Cape Coloureds, 20 Xhosa and 20 Indonesians, combined with HapMap and HGDP samples. They covered 900,0000 SNPs.

Here is a figure where they display the geographical origins of their samples:

You should be familiar with most of these; e.g., the South Asian samples are from Pakistan, from the HGDP, the Europeans are Utah Whites from the HapMap.

Below are pairwise genetic distance values:

I want to emphasize that these are pairwise values. The initials represent:

COL = Coloured
XHO = Xhosa
SAN = Bushmen
YRI = Yoruba
CEU = Utah White
CHB = Chinese (Beijing)
SAS = South Asian (Pakistan)
IND = Indonesian

The relatively close relationship between the Cape Coloureds South Asians should be viewed with caution, since there is evidence that South Asians themselves are a compound admixture between quasi-Europeans and an ancient East Eurasian population of Indians. Note the small distance between South Asians and Europeans in the table above.

An easy way to visualize relationships are principal component scatter plots. These plots show independent dimensions of variation within data, in this case genetic data. The points are individuals, and in 2-dimensional plots generally the x-axis illustrates the first component of variation while the y-axis the second component. These are respectively the independent dimensions which explain the most variation in the data set.

In this paper they constructed a series of PC scatter plots where the axes display between group variation, upon which later groups are mapped upon. Below I've repurposed figure 2 a bit to fit on the screen. Here is the legend:

Figure 2. Principal components scatter plots. Indicating (A) our Coloured and isiXhosa samples across an African versus non-African axis, (B) a Bantu versus Bushmen axis, (C) European versus Asian (Han Chinese, Indonesian and South Asian) axis, (D) Chinese versus Indonesian axis and (E) European versus South
Asian axis.

In other words, what you're seeing are the Cape Coloureds being projected onto PC scatter plots where the axes are constructed from populations which resemble their putative parental populations.

And finally, bringing it all together:

The scatter of Cape Coloureds in the domains between these populations dovetails well with what intuition would suggest if they were in fact an admixture between all of these continental populations. But one issue is that because South Asians and Europeans are relatively close on a continental scale of genetic variation PC plots can only be somewhat definitive in establishing a contribution from that group, though figure 2E above comes close (showing that in relation to South Asian groups Cape Coloureds are shifted toward white Europeans).

To confirm a possible contribution from South Asians they used a regression method whereby the allele frequencies in the admixed population can be modeled as a combination of the frequencies from the non-admixed populations. Table 4 shows the results:

These align well with earlier results. But again there is an issue with the fact that South Asians and Europeans exhibit relatively small between population difference in gene frequencies (the methods rely in part on taking the difference between allele frequencies). They divided the Coloured genome into blocks and used a model where they included South Asians and left them out; the model with South Asians had much greater predictive value than that without South Asians. As I noted above the South Asians who were ancestral to the Cape Coloureds were likely Bengalis and South Indians, not the Pakistanis in the HDGP. Therefore using the model of some of the researchers in this paper the ancestors probably had an "Ancestral North Indian" (quasi-European ANI) component closer to 50% as against 70-80% in the case of the South Asian populations in the HGDP. Though even the ANI was at some genetic distance from modern Europeans, the fact that South Asians themselves can be modeled as an admixture of two populations, one very similar to Europeans, and one more similar to Indonesians than Europeans ("Ancestral South Indians", ASI), though still at a much further remove from Indonesians than the ANI from Europeans, I think explains the relatively modest Fst values between South Asians and Cape Coloureds. The Cape Coloureds in to some extent recapitulate an aspect of an the admixture event which produced South Asians (ANI:ASI::European:Indonesian).

One final dynamic confirmed by these results is that there was "female mediated Asian gene flow" and "male mediated European gene flow" in the emergence of a mixed-race group in South Africa. It is especially striking in the case of Indonesian ancestry, and to me surprisingly small in the case of the African ancestry. In the text the authors note that many marriages were contracted between whites and freeborn Africans, so what we may be seeing here is simply that sociological reality, prior to the further crystallization of a racial caste system African males were able to join the mixed-race community. By contrast Asians, disproportionately slaves, could not marry free whites. This resulted in much greater asymmetrical gene flow between white owners and female slaves.

I'll let the authors conclude:

In conclusion, using a novel method for computing degree of admixture, we demonstrate clear evidence of African (genetically close to isiXhosa), Indonesian, European and South Asian contributions in our recently ( 350 years) admixed Coloured samples. This agrees well within the historical evidence. Therefore the Coloured people represent a new class of unique genomes created from a divergent genetic background, including more than one of the described six major ancestral human genetic clusters...This admixture holds strong potential to offer new insights into complex gene-gene and gene-environment interactions, insight into human evolution and human disease evolution, and enabling medical research efforts unparalleled by any other population.

To those who assume that racial mixture leads to an indistinct and homogeneous mass, here is an extended family of Cape Coloureds:

One issue that a researcher on this paper pointed out to me is that there is a difference between the mix of populations ideally used to predict allele frequencies in populations, and the actually admixing proportions. This makes some sense when you note the difficulties involved in separating South Asian from European contributions because the two groups are the relatively similar in relation to the others. The Fst between the Bushmen and Xhoas is 3 times greater than between the South Asians and Europeans!

A final thought. We don't need to go South Africa for a very admixed population. How about Hawaii? At least 20% of the population in Hawaii apparently exhibits some sort of mixed ancestry (probably an underestimate because of various political issues related to Native Hawaiians).

Citation: Nick Patterson , Desiree C. Petersen , Richard E. van der Ross , Herawati Sudoyo , Richard H. Glashoff , Sangkot Marzuki , David Reich , and Vanessa M. Hayes
Genetic structure of a unique admixed population: implications for medical research
Human Molecular Genetics Advance Access published on November 18, 2009, DOI 10.1093/hmg/ddp505.

razib Sat, 12/05/2009 - 14:53

Tags

Population Substructure

race

south africa

Population substructure within China

razib — Wed, 25 Nov 2009 16:35:43 +0000

Population substructure within China

The state of China has 1/5 of humanity within its borders, so it's genetic structure is of interest. It is obviously important for medical reasons to clarify issues of population structure so that disease susceptibility among the Han is well characterized, in particular with the heightened medical needs of an aging population in the coming generation. And of course, there are the nationalistic concerns. About 20 years ago L. L. Cavalli-Sforza reported that his South Chinese samples were genetically closer to Southeast Asians than North Chinese in The History and Geography of Human Genes. This result has been somewhat muddled in the past generation with the rise of uniparental markers (NRY and mtDNA passed through the male and female lineages) along with studies which utilize hundreds of thousands of SNPs. One thing that seems to be clear is that genes vary as a function of geography in China (just as they do pretty much everywhere).

Two new articles in AJHG shed some more light on this issue, Genomic Dissection of Population Substructure of Han Chinese and Its Implication in Association Studies:

To date, most genome-wide association studies (GWAS) and studies of fine-scale population structure have been conducted primarily on Europeans. Han Chinese, the largest ethnic group in the world, composing 20% of the entire global human population, is largely underrepresented in such studies. A well-recognized challenge is the fact that population structure can cause spurious associations in GWAS. In this study, we examined population substructures in a diverse set of over 1700 Han Chinese samples collected from 26 regions across China, each genotyped at â¼160K single-nucleotide polymorphisms (SNPs). Our results showed that the Han Chinese population is intricately substructured, with the main observed clusters corresponding roughly to northern Han, central Han, and southern Han. However, simulated case-control studies showed that genetic differentiation among these clusters, although very small (FST = 0.0002 â¼0.0009), is sufficient to lead to an inflated rate of false-positive results even when the sample size is moderate. The top two SNPs with the greatest frequency differences between the northern Han and southern Han clusters (FST > 0.06) were found in the FADS2 gene, which associates with the fatty acid composition in phospholipids, and in the HLA complex P5 gene (HCP5), which associates with HIV infection, psoriasis, and psoriatic arthritis. Ingenuity Pathway Analysis (IPA) showed that most differentiated genes among clusters are involved in cardiac arteriopathy (p < 10â101). These signals indicating significant differences among Han Chinese subpopulations should be carefully explained in case they are also detected in association studies, especially when sample sources are diverse.

And, Genetic Structure of the Han Chinese Population Revealed by Genome-wide SNP Variation:

Population stratification is a potential problem for genome-wide association studies (GWAS), confounding results and causing spurious associations. Hence, understanding how allele frequencies vary across geographic regions or among subpopulations is an important prelude to analyzing GWAS data. Using over 350,000 genome-wide autosomal SNPs in over 6000 Han Chinese samples from ten provinces of China, our study revealed a one-dimensional "north-south" population structure and a close correlation between geography and the genetic structure of the Han Chinese. The north-south population structure is consistent with the historical migration pattern of the Han Chinese population. Metropolitan cities in China were, however, more diffused "outliers," probably because of the impact of modern migration of peoples. At a very local scale within the Guangdong province, we observed evidence of population structure among dialect groups, probably on account of endogamy within these dialects. Via simulation, we show that empirical levels of population structure observed across modern China can cause spurious associations in GWAS if not properly handled. In the Han Chinese, geographic matching is a good proxy for genetic matching, particularly in validation and candidate-gene studies in which population stratification cannot be directly accessed and accounted for because of the lack of genome-wide data, with the exception of the metropolitan cities, where geographical location is no longer a good indicator of ancestral origin. Our findings are important for designing GWAS in the Chinese population, an activity that is expected to intensify greatly in the near future.

Below is a PC chart which shows PC1 on the x-axis and PC2 on the y-axis. In green are South Chinese, and in blue the North Chinese. Japanese are the cluster to the top left, and red represents the HapMap Chinese sample.

And here's a visualization of the ancestries of individuals from particular provinces and dialect groups using Structure (right) and Frappe (left) (the K's represent 2 or 3 putative ancestral populations respectively). It's ordered by ancestry within the classes. The rough geographical correlate is north-south. Note the variance in Singapore; most Singaporean Chinese derive from Fujian (with a large Hakka minority, and some Malay admixture on the part of Baba Chinese), but there were enough disparate migratory events that you don't see a bottleneck and decrease in homogeneity compared to Chinese provinces. On the contrary. A minority of Singaporeans seem to be of North Chinese provenance, a result that would not surprising in Taiwan, where such a migration is historically documented (after the fall of Nationalist China), but is more curious in Singapore which was presumably part of the greater Fujianese Diaspora.

Finally, here are pairwise Fst values. Remember that this captures the proportion of genetic variance between populations. Fst values between continental races is on the order of 0.15. This means 15% of the genetic variation is between races. The values below seem to show a maximum between province/dialect difference in China of about 0.5% of the genetic variation. But despite this small value, note how obvious it is above to differentiate individuals from northern and southern regions of China.

Here are some comparable Fst values from Europe:

0.001 = Bulgaria-Austria
0.002 = Poland-Sweden
0.003 = Northern Italy-Switzerland
0.004 = Spain-Sweden
0.005 = Russia-France

I've left out the highest Fst values in Europe, which are between Finns and Southern Italians, on the order of 0.015. But from these data it looks as if Han Chinese are in the same order of magnitude of variance as Europeans in terms of their genetics, but a factor or two lower. But it may be that the coverage of genetic variation is just not as thick in China so that outlier Han populations, the equivalent of Finns (perhaps Sinicized groups in Yunnan?), are out there waiting to push the mean variance higher. It is interesting, though not totally surprising, that different dialect groups in the same region exhibit large genetic differences. Language & genes often correlate because the former circumscribes the limits of marriage networks. The Teochew migrated from Fujian to Guangdong (to my knowledge they are the dominant Chinese group in Thailand), and are nearly as genetically distant from their Cantonese speaking neighbors as they are from North Chinese. Interestingly, the Hakka group who are derived from North Chinese migrants according their history, seem to be closer to "indigenous" South Chinese. Nevertheless, they exhibit less genetic difference from North Chinese than do Cantonese speakers in Guangdong. This is obviously the tip of the iceberg, I suspect that the genetic topography of South China in particular will be surprising because of its geographical fragmentation, the role of powerful clan networks, and the recurrent history of migration from the North China plain by groups who manage to maintain their identities (.e.g, Hakka).*

Citation: Jieming Chen, Houfeng Zheng, Jin-Xin Bei, Liangdan Sun, Wei-hua Jia, Tao Li, Furen Zhang, Mark Seielstad, Yi-Xin Zeng, Xuejun Zhang, and Jianjun Liu, Genetic Structure of the Han Chinese Population Revealed by Genome-wide SNP Variation, doi:10.1016/j.ajhg.2009.10.016

Citation: Shuhua Xu, Xianyong Yin, Shilin Li, Wenfei Jin, Haiyi Lou, Ling Yang, Xiaohong Gong, Hongyan Wang, Yiping Shen, Xuedong Pan, Yungang He, Yajun Yang, Yi Wang, Wenqing Fu, Yu An, Jiucun Wang, Jingze Tan, Ji Qian, Xiaoli Chen, Xin Zhang, Yangfei Sun, Xuejun Zhang, Bailin Wu, and Li Jin, Genomic Dissection of Population Substructure of Han Chinese and Its Implication in Association Studies, doi:10.1016/j.ajhg.2009.10.015

* It is attested that many groups emigrated from South China to North China, but it seems to me that these groups were simply absorbed. I suspect it has to do with the flat topography of the North China plain which does not allow for easy separation between groups. In South China the Hakka tended to farm the more marginal lands, in particular upland regions.

razib Wed, 11/25/2009 - 11:35

Tags

Population Substructure

Some years ago on GNXP Classic, I related what I had been told by a distinguished Chinese medical specialist in Hong Kong, that all of the cases of naso-pharyngeal cancer that he had seen had been in Cantonese patients. He said "...in southern Chinese, in fact in Cantonese." I said "So it's genetic" and he said "Yes, it's genetic." He was clearly in no doubt that Cantonese were not just a dialect and cultural group, but that native Cantonese-speaking people were genetically differentiated from other Han, even other southern Han, based on his practical observations of disease susceptibility.

In southern China today you can still see walled villages populated by e.g. Hakka, although this physical separation/stratification has rapidly begun to disappear within the past generation or so.

So does this mean that there is no Chinese race now, at least in the European ethno-national sense? It seems the Chinese achieved what neither Caesar, Charlemagne, Bonaparte, or Brussels could. Forging one nation out of a group of disparate (but closely related) peoples.

So does this mean that there is no Chinese race now, at least in the European ethno-national sense?

well, i don't see as that surprising, do you? china's about the scale of continent europe. one important implication of this might be to be a bit cautious about extrapolating samples drawn from particular urban areas to all chinese for medical purposes. american chinese tend to be cantonese and fujianese, so american studies based on chinese americans using our citizens as proxies for chinese are going to have an automatic regional bias.

It seems the Chinese achieved what neither Caesar, Charlemagne, Bonaparte, or Brussels could. Forging one nation out of a group of disparate (but closely related) peoples.

i think to some extent this is wrong in the case of caesar. the peoples of the roman empire did view themselvse as romans, and their land as romania. the latinized-descendants of gauls in what became france are referred to as "romans" in the literature of the frankish period (similar in spain). what happened is that unlike china europe remained disaggregated after the collapse of rome.

There has been a remarkable continuity of Chinese history, culture and written language, but genes vary with geography at every scale. That is not to say that you cannot distinguish genetically a Chinese population from a population in a different area, just that there is stratification within the population.

http://scienceblogs.com/gnxp/2008/12/genetic_map_of_east_asia.php

http://scienceblogs.com/gnxp/2009/05/genetic_relations_of_different.php

When the papers refer to different "dialects", they are referring to differences of a type that elsewhere in the world would lead to reference to different "languages". Do I have that right?

yes, dialects are equivalent to languages. or in some cases, as in serbian and croatian, dialects are called languages :-)

bio, I'm not buying into that argument. I will just observe that mutually unintelligible speech is a barrier to exogamy. Except for invading armies, obviously.

Sex differences in ancestry in the New World

razib — Wed, 25 Nov 2009 01:44:19 +0000

Sex differences in ancestry in the New World

A new paper in PLoS ONE, Evaluation of Group Genetic Ancestry of Populations from Philadelphia and Dakar in the Context of Sex-Biased Admixture in the Americas, doesn't add much to what we know. They looked at a several hundred individuals who are self-identified as African American and European American, as well as 49 Senegalese from Dakar. Additionally, they reanalyzed data from Latin America from whites and blacks in Brazil, as well as a group of mixed Cubans. They found what you might expect to find, African and Native ancestry shows a female bias, European ancestry shows a male bias.

But Figure 3 is nice in that it illustrates how exceptional European Americans are: they are a New World population with very little admixture, that is, a "pure race." The data on people of white identification in most of Latin America, including in the European dominated southern cone, is that a non-trivial load of non-European genes exists in these populations (often indigenous ancestry down the maternal lineage). Similarly, African origin populations in the New World are invariably known to be admixed, more or less. Finally, even the "indigenous" generally show non-trivial European or African ancestry (this is evident in the regular notations of African or European ancestry being used to explain anomalous results).*

Citation: Stefflova K, Dulik MC, Pai AA, Walker AH, Zeigler-Johnson CM, et al. 2009 Evaluation of Group Genetic Ancestry of Populations from Philadelphia and Dakar in the Context of Sex-Biased Admixture in the Americas. PLoS ONE 4(11): e7842. doi:10.1371/journal.pone.0007842

* Some escaped African slaves "went native" and assimilated into indigenous tribes. The symbiotic relationship between the Seminoles of Florida and escaped blacks is well known, but not exceptional. The case in the United States is exceptional in its magnitude in that most people of native self-identifications are mixed.

razib Tue, 11/24/2009 - 20:44

Tags

Population Substructure

race

White Americans from Philadelphia are unlikely to have any Native or African ancestry. Most are descended primarily from 19th or early 20th century Irish, Polish, Italian and Jewish immigrants.
Of course the amount of Native ancestry in Euro-Americans has been exaggerated by family legends, and the actual amount, where it exists, is often going to be undetectable by commercially available DNA tests.
If one is looking for such ancestry, it is best to look at French Canadians, as well as old-stock Americans with Scots and Scots-Irish roots, especially in the South and West. Scots were heavily involved in the colonial-era southeastern deerskin trade (likewise the later Canadian fur trade) and many of the early mountain men were of Scots-Irish descent.

The Euro-American results are false.
http://www.plosone.org/article/slideshow.action?uri=info:doi/10.1371/jo…
No Negroid mtDNA was found in the Euro-American sample. The only African mtDNA is two North African Caucasoid lineages. These are most likely from Southern Europeans.
None of the "Native American/Asian" lineages are Amerind.
http://www.plosone.org/article/slideshow.action?uri=info:doi/10.1371/jo…

Not surprising that a small proportion of North African mtDNA might be found in Euro-Americans from Philadelphia, given the number of Philadelphians of Southern Italian/Sicilian descent.
Among Euro-Americans in general, there is clearly less non-European ancestry than among the most "European" populations in Latin America (Argentines) or even among white Afrikaners (many of whom have minor African or South/Southeast Asian ancestry). But if one is looking for such ancestry among Euro-Americans, the sample used in this case is about the LAST place one would want to look.

Inferring deep history in genes

razib — Mon, 26 Oct 2009 02:48:17 +0000

Inferring deep history in genes

When Mendelism reemerged in the early 20th century to become what we term genetics no doubt the early practitioners of the nascent field would have been surprised to see where it went. The centrality of of DNA as the substrate which encodes genetic information in the 1950s opened up molecular biology and led to the biophysical strain which remains prominent in genetics. Later, in the 1970s Alan Wilson and Vincent Sarich used crude measures of genetic distance to resolve controversies in paleontology, specifically, the date of separation between the human and ape lineage. Genetics spans the physical and historical sciences, whereas physically oriented scientists may look to DNA as a basis for computation, historically oriented scholars can use it to illuminate mysteries in their own fields.

In the 1980s the "mitochondrial Eve" arrived on the scene, purporting to map out the demographic history of our species over the past 200,000 years. This was during an era when extraction and amplification of genetic material was primitive, and so the numerous mitochondria were the preferred sources of information. Additionally, the uniparental nature of mtDNA makes it ideal for a coalescent model.

Over the past two decades science has come much farther. Genetic material is easier to analyze, and the computers to do that analysis have become much more powerful. A non-trivial segment of the genome is now being brought to bear on questions of genetic history. More powerful computational techniques mean that the complexity of the models can be cranked up. This is evident in a recent paper, Inferring the Joint Demographic History of Multiple Populations from Multidimensional SNP Frequency Data:

Demographic models built from genetic data play important roles in illuminating prehistorical events and serving as null models in genome scans for selection. We introduce an inference method based on the joint frequency spectrum of genetic variants within and between populations. For candidate models we numerically compute the expected spectrum using a diffusion approximation to the one-locus, two-allele Wright-Fisher process, involving up to three simultaneous populations. Our approach is a composite likelihood scheme, since linkage between neutral loci alters the variance but not the expectation of the frequency spectrum. We thus use bootstraps incorporating linkage to estimate uncertainties for parameters and significance values for hypothesis tests. Our method can also incorporate selection on single sites, predicting the joint distribution of selected alleles among populations experiencing a bevy of evolutionary forces, including expansions, contractions, migrations, and admixture. We model human expansion out of Africa and the settlement of the New World, using 5 Mb of noncoding DNA resequenced in 68 individuals from 4 populations (YRI, CHB, CEU, and MXL) by the Environmental Genome Project. We infer divergence between West African and Eurasian populations 140 thousand years ago (95% confidence interval: 40-270 kya). This is earlier than other genetic studies, in part because we incorporate migration. We estimate the European (CEU) and East Asian (CHB) divergence time to be 23 kya (95% c.i.: 17-43 kya), long after archeological evidence places modern humans in Europe. Finally, we estimate divergence between East Asians (CHB) and Mexican-Americans (MXL) of 22 kya (95% c.i.: 16.3-26.9 kya), and our analysis yields no evidence for subsequent migration. Furthermore, combining our demographic model with a previously estimated distribution of selective effects among newly arising amino acid mutations accurately predicts the frequency spectrum of nonsynonymous variants across three continental populations (YRI, CHB, CEU).

The author summary is a bit plainer:

The demographic history of our species is reflected in patterns of genetic variation within and among populations. We developed an efficient method for calculating the expected distribution of genetic variation, given a demographic model including such events as population size changes, population splits and joins, and migration. We applied our approach to publicly available human sequencing data, searching for models that best reproduce the observed patterns. Our joint analysis of data from African, European, and Asian populations yielded new dates for when these populations diverged. In particular, we found that African and Eurasian populations diverged around 100,000 years ago. This is earlier than other genetic studies suggest, because our model includes the effects of migration, which we found to be important for reproducing observed patterns of variation in the data. We also analyzed data from European, Asian, and Mexican populations to model the peopling of the Americas. Here, we find no evidence for recurrent migration after East Asian and Native American populations diverged. Our methods are not limited to studying humans, and we hope that future sequencing projects will offer more insights into the history of both our own species and others

The basic building block of this paper is "allele frequency spectrum," or AFS. The text explains it rather well:

Given a genetic region sequenced in multiple individuals from each of P populations, the resulting AFS is a P-dimensional matrix. Each entry of this matrix records the number of diallelic genetic polymorphisms in which the derived allele was found in the corresponding number of samples from each population. For example, if diploid individuals from two populations were sequenced, with 10 individuals from population 1 and 5 from population 2, the AFS would be a 21-by-11 matrix (indexed from 0). The [2,0] entry would record the number of polymorphisms for which the derived allele was seen twice in population 1 but never seen in population 2, while the [20],[5] entry would record polymorphisms for which the derived allele was homozygous in all individuals from population 1 and seen 5 times in population 2. If all polymorphic sites possess only two alleles and can be considered independent, the AFS is a complete summary of the data. Many of the statistics commonly used for population genetic inference, such as F_ST and Tajima's D , are summaries of the AFS

The model also uses diffusion equations to project change in genetic frequencies, a technique first utilized by R. A. Fisher. Since Fisher's previous research had been in thermodynamics, it was natural for him to analogize change in gene frequencies to the sort of fluxes which one sees in heat related processes. It seems that the combination of approaches that they used here is moderately scalable:

The computational advantage of the diffusion method is even larger when placed in the context of parameter optimization. Unlike the coalescent approach, there is no simulation variance, so efficient derivative-based optimization methods can be used. As examples, consider our applications to human data, which involve 20 samples per population. On a modern workstation, fitting a single-population three-parameter model took roughly a minute, while fitting a two-population six-parameter model took roughly 10 minutes. The fits of three-population models with roughly a dozen parameters typically took a few hours to converge from a reasonable initial parameter set. This speed allows us to use extensive bootstrapping to estimate variances, overcoming the limitations of composite likelihood.

They did have to stop at 3 populations because of computational problems when attempting to analyze 4 populations. When looking at the Mexican American data set they discarded Africans for the purpose of comparison with outgroups. It seems the biggest step forward here is that they are able to add in a host of evolutionary genetic parameters into their model, in particular selection and migration, without too much difficulty. The technique also seems robust to the deviations which selection or linkage would introduce into evolutionary history of populations where neutral evolution would be the markers from which one would infer population fission. Previous models of population movements out of Africa were often unrealistic insofar as it was assumed that populations were exclusively fissiparous and that future back-migration would never occur. That is fine as it goes, but if it is possible to construct models which don't explode because of computational intensity then it seems prudent to do so.

Reading the whole paper through it seems that this is more a signpost to the future in method rather than a treasure-trove of results. Shifting particular demographic events by a few tens of thousands of years is important on the margin, but with the confidence intervals I'm not sure what to make of the results on their own (though in concert with other findings they are of more interest). Instead, consider that they've put the software which they used to generate these results online:

Diffusion Approximation for Demographic Inference (âaâi )

Citation: Gutenkunst RN, Hernandez RD, Williamson SH, Bustamante CD, 2009 Inferring the Joint Demographic History of Multiple Populations from Multidimensional SNP Frequency Data. PLoS Genet 5(10): e1000695. doi:10.1371/journal.pgen.1000695

razib Sun, 10/25/2009 - 22:48

Tags

biology

genetics

Population Genetics

Population Substructure

The mostly South Asian origins of Indian Muslims

razib — Sat, 10 Oct 2009 17:07:21 +0000

The mostly South Asian origins of Indian Muslims

Dienekes points to a new paper which attempts to quantify the genetic ancestry of South Asian Muslims into indigenous and exogenous components:

Islam is the second most practiced religion in India, next to Hinduism. It is still unclear whether the spread of Islam in India has been only a cultural transformation or is associated with detectable levels of gene flow. To estimate the contribution of West Asian and Arabian admixture to Indian Muslims, we assessed genetic variation in mtDNA, Y-chromosomal and LCT/MCM6 markers in 472, 431 and 476 samples, respectively, representing six Muslim communities from different geographical regions of India. We found that most of the Indian Muslim populations received their major genetic input from geographically close non-Muslim populations. However, low levels of likely sub-Saharan African, Arabian and West Asian admixture were also observed among Indian Muslims in the form of L0a2a2 mtDNA and E1b1b1a and J*(xJ2) Y-chromosomal lineages. The distinction between Iranian and Arabian sources was difficult to make with mtDNA and the Y chromosome, as the estimates were highly correlated because of similar gene pool compositions in the sources. In contrast, the LCT/MCM6 locus, which shows a clear distinction between the two sources, enabled us to rule out significant gene flow from Arabia. Overall, our results support a model according to which the spread of Islam in India was predominantly cultural conversion associated with minor but still detectable levels of gene flow from outside, primarily from Iran and Central Asia, rather than directly from the Arabian Peninsula.

This is entirely predictable; there's a reason that during communal riots men are forced to expose their penises, aside from dress distinguishing a Muslim from a Hindu is not really possible. A small number of very high status Muslims are obviously predominantly non-South Asian, they look different. This stratum still intermarries internationally, e.g., Benazir Bhutto's mother was Kurdish (one reason she looked so white). When the British first arrived in the Indian subcontinent they distinguished between white and black Muslims just as the local elite Muslims did.

Nevertheless, there was a great deal of intermarriage. Even at the commanding heights of the white Muslim aristocracy, the Mughals, there was outmarriage. Shah Jahan was 3/4 Rajput. Compare Shah Jahan to his grandfather, Akbar (a contemporary portrait).* The Mughals were not the best record keepers (they were a rentier state par excellence), but the composition of their military and civil service is known because of differential grants of income to categories of service elites. Foreign Muslims, those born outside India, were at a premium, and often comprised a significant proportion every generation (Persians being siphoned to the civilian bureaucracy, Turks and Pathans to the military). This is in keeping with a tendency among the Turkic Muslim dynasties in general, from the Mamlukes to the Ottomans, who replenished their non-hereditary elite from without. Where did the offspring of these newcomers go? Likely they descended down the class hierarchy, intermarrying with local Muslim converts (or, even facilitating the conversion of a high status Hindu family through intermarriage), who would be eager to acquire Persian or Turkic ancestry for their descendants and so a higher heritable status.

But these data do show an exception. The Bohra. Unlike the other Muslim groups, there is basically no evidence of significant (i.e., above expectation) Middle Eastern lineages among them. Why? You need to know a little history and ethnography. The Bohras are an Ismaili group, a minority sect within Shiism, and very marginal due to their peculiar history and theology. Recent works of religious history suggest that Ismailis were extremely influential in spreading Islamic ideas among the populations of South Asia. These researchers suggest that Ismaili Islam served as a "gateway" to Sunni Islam for many South Asians. Additionally, it may be that many Hindu groups (Hussaini Brahmins and the Megh for example) were influenced by this variant of Islam. The Ismaili flavor of South Asian Islam was eclipsed by the rise of the institutional structure of Sunni orthodoxy (which included the state, as well as Sufi orders), most prominently among the late Mughals such as Aurangzeb. It was during this period when there were massive shifts from Ismailism to Sunni Islam, and some suggestions that many Hindu groups moved away from syncretism with Islam lest they be labelled heretics by the state, a far worse position than simply being heathen.

The Bohras are one group of Ismailis who did not convert to Sunni Islam. Some scholars assert that before the Mughal period the majority of Muslims in Gujarat were Ismaili, while after only a minority were. Like South Asian Muslims in general the origin of Ismailism was among converts from what became Hinduism, in particular the trader castes in the instance of the Bohras. These data give tentative support to that contention, the mix of lineages is what is common among Indian upper castes. I speculate that there is a possibility that there was selective conversion to Sunni Islam, and that those from mercantile backgrounds were the least incentivized to switch to the "orthodox" confession. In contrast, those reliant on state beneficence, or those who had Sunni landlords, were probably more likely to switch. There is some suggestion that the Parsis, who were the remnants of Iranian Zoroastrianism which emigrated to India, were mostly from priestly families. This group obviously had least incentive to convert to Islam because many of their privileges were bundled up with a Zoroastrian religious identity. In any case, this can be tied back up to the intermarriage of Iranian and Central Asian Muslims with local Muslim elites: there would be no incentive for non-South Asian Muslims to marry into Ismaili families who were on the wrong side of the Mughal state which had served as their initial patrons. Rather, it would be contingent upon any native Muslim family who wished to intermarry with a more prestigious Persian or Central Asian lineage to toe the line of religious orthodoxy, whether that be Sunnism, or, the less heterodox Twelver Shiism (there were Twelver Shia polities in India).

An alternative hypothesis is that the scholars who claim many more Ismailis in the past are misreading the textual evidence, and that Ismailism was a much smaller sect than they envisage. That is never went much beyond the trader castes, and so the possibility for intermarriage with prestigious non-South Asian lineages was already minimized because of that group's preference for wealth based on land rents.

Note: Though the authors dismiss this, I suspect that Dienekes' is correct in observing that the Mapplas of Kerala do have Arab ancestry, which other South Asian Muslims do not. This is culturally expected, the Mappla form of Islam owes much more to South Arabian exemplars than Sunni Islam in the rest of the subcontinent, which is Turkic or Persian in flavor.

Citation: Traces of sub-Saharan and Middle Eastern lineages in Indian Muslim populations, European Journal of Human Genetics, doi: 10.1038/ejhg.2009.168

* The portraits may be idealizations, but from what I have read Akbar did look much more Mongolian than his descendants. Aurangzeb, Shah Jahan's half-Persian son, was noted to have particularly fair skin, likely in contrast with his father who was mostly South Asian in ancestry.

razib Sat, 10/10/2009 - 13:07

Tags

Population Substructure

I met a rich blonde lady in Turkey whose ancestry was from this international Muslim elite you speak of. One quarter of her ancestors were Muslims from up in the middle of Russia, east of Moscow.

I met a rich blonde lady in Turkey whose ancestry was from this international Muslim elite you speak of. One quarter of her ancestors were Muslims from up in the middle of Russia, east of Moscow.

much of what is now mostly orthodox (notionally) and russian in the ural/volga region and east was once muslim and turkic. after the conquests of these regions by the russians in the 17th and 18th centuries there were three paths:

1) become russian and orthodox. many boyars came from turkic and muslim families by origin.

2) figure out a way to remain distinctive religiously & ethnically in the context of residing within a polity dominated by russians and dedicated to orthodoxy.

3) emigrate, especially to the ottoman empire, since one line of muslim though argues that muslims should never live under the rule of non-muslims.

during the reign of catharine the great there were mass migrations to the other side of the black sea as the russian empire finally swallowed up the remaining turkic muslim polities in russia proper.

more generally, islam was an international civilization from almost its beginning. the umayyad caliphs had control (at least more than nominal) from spain to southern pakistan. after the collapse of the political order, civilizational unity through common norms meant that travelers like ibn battuta could move between africa and asia while remaining in a muslim world.

"low levels of likely sub-Saharan African ... admixture were also observed": is that likely to be from the Islamic slave trade across the Indian Ocean, or is it impossible to tell?

The interesting question is the manner and rate of conversion. At what date did the Arabian empire have a majority Muslim population, rather than just a Muslim ruing class? The Jizya (and the desire to escape paying it) was probably a major factor in conversion in the Caliphate. In India, I've often read that it was lower caste Hindus who converted, because Hinduism wasn't doing them any favours. But do we know if this is really true?

Russian religious history was especially turbulent, with lots of heretics, judaizing Christians, christianizing Jews, and probably the same with Muslims.

Islam converts people tax structure -- the Jizyah. Islam is just more economical for a given individual.

The power to tax is the power to destroy. I imagine that the libertarians have already run with this one.

It is not as if mostly low castes Hindus converted to Islam.
Since high caste Hindus make a small roportion of the population less than 25%, so obviously more numbers would come from lower castes. The first mass conversion in India was done by a Hindu Raja in Bulandsahar or Baran-Sahar who was high caste and converted with 20,000 of his peoples. And it was his own voluntarily step

To georgesdelatour :

Jizya is only a financial tribute / aid to the Muslim State which is in-charge of safeguarding the security of the state and non-muslim's lives and properties on their behalf.

Non-muslims pay Jizya BUT they are EXEMPTED from any other taxes which muslims pay in a Muslim State i.e. Zakat, Khums etc.

As compared to taxes which the Muslims are subjected in a Muslim state, the amount of Jizya is very low.

As such, Jizya should not be interpreted as "Additional Tax" imposed on non-muslims. It is rather a "lesser" obligation as compared to that of a Muslim.