ChromoPainter/fineSTRUCTURE analysis of Italy/Balkans/Anatolia

This was done on the same dataset as the previous fastIBD analysis.

The population assignments:

The heatmap, showing relationship between inferred populations:

The principal components analysis:

The correspondence between inferred populations and K12b components:

Results for Project participants can be found in this spreadsheet; remember than in the chunkcounts tabs, columns represent donor and rows recipient populations.

fastIBD analysis of Balkans/West Asia

Now that I've discovered a way to boost Clusters Galore analysis even further by using fastIBD, I will start experimenting with different regional populations. This analysis took about 5 hours to complete, so it appears to be quite practical.

For my first experiment, I carry out an analysis of various populations from the Balkans and West Asia.

Clusters Galore

27 different clusters were inferred with 17 MDS dimensions. Some interesting findings:

• For the first time there emerge a couple of clusters that appear to be quite specific to Armenians (#2 and #3). 
• Similarly, Assyrians are broken to a few clusters that appear fairly specific to them  (#9-11)
• Georgians are split into three clusters, one of which (#14) is linked with the neighboring Abkhasians, who in turn have their own exclusive cluster (#25)
• The cluster modal in Greeks (#6) includes 14 of 19 Greek participants, and a few Greeks are also in the Balkan cluster (#8) and an Iranian-Turkish cluster (#4)
• The Behar Cypriot sample also splits into two, and the few Turkish Cypriot participants link to one of them (#13)
• The Ossetian project participant links to one of the three North_Ossetian clusters
• The major Balkan cluster (#8) still defies resolution. I am certain, however, that structure in this cluster will be uncovered with more participation. MCLUST adapts the cluster size and shape, and a "big", inclusive cluster spanning the Balkans appears more parsimonious than smaller clusters centered on the different groups. With larger participation, I anticipate that regional structure will be uncovered in the Balkans as well.

I cannot stress the importance of participation strongly enough. When groups have more participants, it is possible to both:

1. Discover group-specific clusters, by identifying what is common between members of groups
2. Discover within-group clusters, by identifying what is different between members of groups

For example, the great participation of Armenians in the Project has now allowed me to discover structure within the Armenian population. It appears, that cluster #2 corresponds to a more "western" Armenian group, and #3 to a more "eastern" one, with some overlap between the two.

Inter-population IBD


You can also see a visual representation of inter-population IBD:

I have only included populations with 5+ participants in this representation. Reddish shades express high IBD sharing; bluish ones low one. The heatmap has been scaled by row.

As you might expect, values across the diagonal are "reddish", since individuals within populations tend to have high IBD sharing with each other.

A few features "pop out" of the screen. Going from top to bottom:

• Intra-Iranic sharing
• Intra-Armenian sharing
• Intra-Balkan sharing
• Georgian-Abkhaz sharing

You can probably get more out of the figure, but these appear to be the most salient features.

Results for Project Participants


The results can be found in the spreadsheet, and include:

• Probabilities of assignment in each of the 27 clusters of the Clusters Galore analysis
• Z-scores of IBD between each individual and each of the 20 populations with 5+ participants. Higher values mean more IBD sharing. Note that Z-scores have been calculated for each row, hence each participant must scan his own row to find populations with an excess (+) or deficiency (-) of IBD sharing, and people should not compare across different rows.

Last but not least, I want to remind new project participants to leave a message in the Information about Project samples thread. Your comment will not appear immediately, since comment moderation is on, and also note that there are multiple pages of comments. 


If you haven't joined the Project yet, I encourage you to do so if you are eligible.

Yunusbayev et al. (2011) data assessed with Dodecad v3

I have acquired the data from the recent Yunusbayev et al. (2011) paper on the Caucasus. This includes the following populations:

• Kurds_Y 6
• Bulgarians_Y 13
• Ukranians_Y 20
• Mordovians_Y 15
• Armenians_Y 16
• Abhkasians_Y 20
• Balkars_Y 19
• North_Ossetians_Y 15
• Chechens_Y 20
• Nogais_Y 16
• Kumyks_Y 14
• Turkmens_Y 15
• Tajiks_Y 15

It is a valuable new addition to the Project, and it is commendable that it has been made publicly and easily available so swiftly after the appearance of the Yunusbayev et al. (2011) paper.

To get the ball rolling on the new Yunusbayev et al. data, I will map the new populations onto the Dodecad v3 components; they will be added to the Dodecad v3 spreadsheet as they are calculated.

I have been laboriously designing a new global (including Amerindians and Australasians) Dodecad X1 experimental calculator with 3,010 individuals for a few weeks now, but I guess I will now have to reboot it with 3,214.

Together with some other new data I recently discovered, I now have 9,799 individuals (some duplicates from different sources) in my global database. My Dodecad dataset of 511 individuals from a single country or ethnic group isn't too shabby either. Let's hope for a new data release that will push the data collection above the magic 10,000.

UDDATE:

I have added the first 7 populations to the spreadsheet; the others are being calculated as we speak. Most of them seem in line with expectations, but the Abkhasian sample has one outlier individual (abh27), and has thus been placed in the "Outliers" tab of the spreadsheet; a new set of admixture proportions, minus that outlier individual, will be calculated anew:

UPDATE II: The population portraits have been uploaded to Google Docs as a rar file (Sendspace mirror). Average admixture results have all been entered to the spreadsheet.

'bat' calculator (Balkans-Anatolia-Turkic)

I have decided to make a new calculator for DIYDodecad that may be useful for individuals from the Balkans and Anatolia. You can download it from here at Google Docs (or here from sendspace). The terms of use are the same as for DIYDodecad v 2.0. To run it, you simply extract the contents of the RAR file in your working directory, and type bat.par whenever you typed dv3.par in the instructions.

The reference populations can be seen below. I have included all available Balkan populations, as well as Turks and Armenians. Moreover, I have included all available Turkic populations.

The marker set is the same as used in Dodecad v3. Three components emerge: one centered in the northern Balkans, one in eastern Anatolia, and one present in various proportions among all Turkic populations (see Turkic cline).

The components have been named accordingly, but please note that they do not necessarily reflect recent ancestors. For example, it is a good hypothesis that the Anatolia component was present in the Balkans even in ancient times, so one need not seek a recent Anatolian ancestor to explain its presence in a Balkan individual. Similarly for the Balkans component in Anatolia, which may reflect the diverse Balkan peoples that have settled in Anatolia since the dawn of history, so a present-day inhabitant of Anatolia need not seek a recent Balkan ancestor.

Likewise, the Turkic component is only part of the genetic makeup of the Turkic speakers who arrived in Anatolia, since those probably also carried West Eurasian population elements picked up en route from Siberia to Anatolia.

The way to interpret your results is to see whether you have an excess or deficiency of any component relative to your ethnic group. For example, an Anatolian Greek may have a higher Anatolia/Balkans ratio than a Balkan Greek and likewise for a Balkan vs. Anatolian Turk; the latter may also have a variable Turkic component which will reflect differential Central Asian input.

The Dodecad Oracle v1

Here is a little fun tool that tests the Dodecad v3 admixture proportions of an individual against all the reference populations, but also against the best pairwise combinations of these populations.

You need to install R to use it, and then download the program and double click on the file DodecadOracleV1.RData that can be found within the rar file. You will then be faced with a command prompt where you can enter the following commands:

Examining which populations are available

Just enter

X[,1]

You will see a list of 227 populations. You can use these population IDs in the next section.

Which populations are closest to a particular population?

Enter:

DodecadOracle("British_D")
[,1] [,2]
[1,] "British_D" "0"
[2,] "British_Isles_D" "0.9798"
[3,] "Cornwall_1KG" "1.1533"
[4,] "Kent_1KG" "2.265"
[5,] "Irish_D" "3.7643"
[6,] "Dutch_D" "4.5354"
[7,] "Mixed_Germanic_D" "6.8971"
[8,] "Norwegian_D" "11.3111"
[9,] "Orkney_1KG" "12.4652"
[10,] "Orcadian" "12.8195"

If you want to find e.g., the top-30 populations, rather than just the top-10, enter:

DodecadOracle("British_D", k=30)

Which populations are closer to a particular individual?

Enter the admixture proportions of the individual (from the "Individual results" tab of the spreadsheet) as follows:

DodecadOracle(c(4.6, 16.7, 33.6, 0, 23.2, 0.4, 0.6, 1.6, 0.7, 14.1, 4.5, 0.2))
[,1] [,2]
[1,] "Ashkenazi_D" "3.7908"
[2,] "Ashkenazy_Jews" "4.1473"
[3,] "Morocco_Jews" "6.338"
[4,] "S_Italian_Sicilian_D" "12.5443"
[5,] "Sephardic_Jews" "13.5067"
[6,] "C_Italian_D" "14.4554"
[7,] "Sicilian_D" "14.7469"
[8,] "S_Italian_D" "15.748"
[9,] "Tuscan_X" "15.9981"
[10,] "O_Italian_D" "16.1474"

Once again, you can specify k=30, if you desire the 30 top matching populations instead of the default 10.

Mixed Mode

You use mixed mode by adding mixedmode=T in any of the commands. The program then considers all pairs of populations, and for each one of them calculates the minimum distance to the sample in consideration, and the admixture proportions that produce it; population pairs where the distance to one of the two populations is smaller than to any admixture of the two are ignored.

Example:

DodecadOracle("Pathan",mixedmode=T)
[,1] [,2]
[1,] "Pathan" "0"
[2,] "84.8% Pakistani + 15.2% Urkarah" "1.075"
[3,] "84% Pakistani + 16% Stalskoe" "1.1555"
[4,] "63.9% TN_Brahmin + 36.1% Urkarah" "1.6669"
[5,] "32.4% Urkarah + 67.6% Meghawal" "2.3516"
[6,] "56.3% INS + 43.7% Urkarah" "2.4901"
[7,] "11.5% Adygei + 88.5% Pakistani" "2.6245"
[8,] "82.4% Sindhi + 17.6% Stalskoe" "2.6318"
[9,] "62.9% AP_Brahmin + 37.1% Urkarah" "2.7322"
[10,] "11.2% Lezgins + 88.8% Pakistani" "2.7749"

The mixed mode should be used with caution, and it shows, more than anything else, how similar apparent "mixes" can be achieved by different combinations of ancestry. Nonetheless, it may prove somewhat useful. For example, there is a suggestion in the above results, that Pathans can be viewed as a mix of other South Asian populations and populations from the eastern Caucasus, a suggestion that was arrived at independently by the Project using different methods.

Here is another example:

DodecadOracle("Assyrian_D",mixedmode=T)

[,1] [,2]

[1,] "Assyrian_D" "0"

[2,] "83.9% Armenians_16 + 16.1% Yemen_Jews" "1.7829"

[3,] "89.1% Armenian_D + 10.9% Saudis" "2.1624"

[4,] "84.3% Armenians_16 + 15.7% Saudis" "2.2884"

[5,] "88.9% Armenian_D + 11.1% Yemen_Jews" "2.2983"

[6,] "83.8% Armenian_D + 16.2% Bedouin" "4.1579"

[7,] "72.2% Armenian_D + 27.8% Syrians" "4.1841"

[8,] "23.4% Georgians + 76.6% Iraq_Jews" "4.2418"

[9,] "76.2% Armenians_16 + 23.8% Bedouin" "4.332"

[10,] "61.5% Armenians_16 + 38.5% Syrians" "4.4019"

This reaffirms the close relationship of Assyrians to Armenians that has been noticed in the project and by others, and it also shows that Assyrians differ from Armenians in a Southwestern Asian direction, consistent with their Semitic language.

Or, African Americans:

DodecadOracle("ASW",mixedmode=T)

[,1] [,2]

[1,] "ASW" "0"

[2,] "81.3% Hausa + 18.7% N._European" "2.3891"

[3,] "18.4% Orkney_1KG + 81.6% Hausa" "2.4031"

[4,] "18.5% Argyll_1KG + 81.5% Hausa" "2.4268"

[5,] "18.4% Orcadian + 81.6% Hausa" "2.4657"

[6,] "80.5% Igbo + 19.5% N._European" "2.5031"

[7,] "80.6% Brong + 19.4% N._European" "2.523"

[8,] "18.6% CEU + 81.4% Hausa" "2.5938"

[9,] "19.1% Argyll_1KG + 80.9% Brong" "2.6197"

[10,] "19% Orkney_1KG + 81% Brong" "2.6274"

I don't know that much about the slave trade, but I believe that Ghana was an important part of it?

Another thing to watch, is that some populations tend to have more than one sample available, so they appear to be mixtures of themselves, which is not really very informative, e.g., Spanish_D

DodecadOracle("Spanish_D",mixedmode=T)

[,1] [,2]

[1,] "Spanish_D" "0"

[2,] "7.9% French_Basque + 92.1% IBS" "0.8713"

[3,] "68.9% IBS + 31.1% Spaniards" "1.0377"

[4,] "98.8% IBS + 1.2% Irish_D" "1.2959"

[5,] "1.2% British_Isles_D + 98.8% IBS" "1.3018"

[6,] "1.2% British_D + 98.8% IBS" "1.3019"

[7,] "99% IBS + 1% Norwegian_D" "1.3046"

[8,] "1.2% Cornwall_1KG + 98.8% IBS" "1.3048"

[9,] "98.8% IBS + 1.2% Kent_1KG" "1.3142"

[10,] "2.2% French_D + 97.8% IBS" "1.3179"

To deal with these problems, you must "edit" the X matrix if you want to exclude some populations. For example, if you want to exclude "Spaniards" and "IBS", you must enter:

X <- X[setdiff(1:227,which(X[,1]=="IBS" | X[,1]=="Spaniards")),]

but notice, that you must relaunch the program, if you want to get the original matrix, or alternatively save it like this:

Z<-X

and then retrieve it like this:

X<-Z

Structure in West Asian Indo-European groups (part 2)

I will occasionally revisit old posts such as Structure in West Asian Indo-European groups to take advantage of new population samples from project submitters. This time around, I included our first Kurdish and Azeri participants, and limited the analysis to populations for which I had large numbers of markers (the final total is a ~132k pruned set of markers).

I also included Greeks, Caucasian populations (Georgians and Lezgins), and Levantine Arabs (Syrians-Lebanese) who frame this region from the West, North, and South respectively, as well as Assyrians who are interspersed in West Asia as an ethno-religious minority.

Here are dimensions 1 and 2 of the multidimensional scaling plot:

The two Caucasian groups (South and Northeast Caucasian Georgians and Lezgins respectively) form their own clusters. So do the Iranians and the Syro-Lebanese.

(As always population labels are placed on population averages, and _D denoted Dodecad Project populations)

Curiously, many linguists assert a close relationship of Greek and Armenian within the Indo-European language family. Turks speak an Altaic language due to migration of a numerically small population element, but their pre-Turkish genetic ancestors were Anatolian speakers, Greeks, Armenians, and Iranians, i.e., primarily Indo-Europeans.

Dimensions 1 and 3:

Dimension 3 contrasts Greeks from West Asian groups. Notice also the presence of 3 Armenians at the bottom of the plot, these are outliers of the Behar et al. Armenian sample.

Notice that the Azeri_D sample clustered with Iranians in dimension 2 and with Turks in dimension 3. This is not very surprising, as Azeris speak a Turkic language, but also have clear Iranian antescendants. The Kurd_D sample, on the other hand, clusters consistently with Iranians.

The variability of the Greek_D population sample along dimension 3 is also quite interesting. This could reflect variable levels of influence of extra-Greek European/Anatolian population elements on the basic Greek stock. Greeks who possess 23andMe or FTDNA Illumina population data are strongly encouraged to join the Project to help us better determine regional variation within the ethnic Greek population.

The Clusters Galore analysis results are as follows (9 clusters/3 MDS dimensions retained) :

In brief, the modal populations in each cluster are:

1. Greeks
2. Iranians
3. Turks
4. Syrians and Lebanese
5. Armenians and Georgians
6. Armenians and Assyrians
7. Syrians and Iranians
8. Lezgins
9. Georgians

I will be happy to provide to all Dodecad Project members from the _D populations with their individual results. If you send me e-mail at dodecad@gmail.com, I will send you a line with your probabilities of assignment in each of the 9 clusters, as well as the 3 co-ordinates in the first 3 MDS dimensions plotted above.

I strongly encourage individuals from West Asia, the Balkans, and Italy to contact me for inclusion in the Project (send e-mail first, not data). While submission to the Project is currently closed, I usually accept data from these regions.

ADMIXTURE analysis with Dodecad Populations (update #2)

Thanks to all the participants of the Project, the number of populations has increased, and so have sample sizes within pre-existing populations in the Project. There are now 17 populations with at least 5 individuals in the Project:

Assyrian, Scandinavian, Greek, Finnish, S_Italian_Sicilian, Ashkenazi, German, Indian, Portuguese, Armenian, Russian, Spanish, British, Irish, Turkish, N_Italian, Balkans

Below are the K=10 ADMIXTURE results with these populations:

Admixture proportions can be found in the spreadsheet.

The fact that the addition of 17 populations and 143 individuals to the core set of 36 populations and 692 individuals results in the same 10 ancestral components testifies to the stability of this solution. Hopefully, within 2011 I will develop an even better comparison set to work with.

Another test of the validity of the analysis is comparison of independent samples of the same populations:

Ashkenazi, Armenian, Spanish, Turkish, N_Italian

I have a sample of Dodecad Project members for each of the above, as well as a published population. A way to measure the concordance between the two is to calculate the correlation coefficient (rounded to the 3rd decimal point):

• Ashkenazi Jews: 0.999
• Armenians: 0.988
• Spanish: 0.998
• Turkish: 0.995
• N_Italian: 0.996

The concordance is remarkable.

I have also made a RAR of "population portraits". It is important to do this to determine whether minor ancestral components represent population-wide phenomena or are limited to a few individuals.

For example, here are the Turks of the Dodecad project:

The sample is a bit more varied than the sample included in Behar et al:

This probably underscores the importance of broad coverage of large countries and ethnic groups, as I have discovered recently in my analysis of 9 different populations of Pakistan.

Another new population are the Irish, presenting a picture of remarkable homogeneity:

Here is the population portrait for the Balkans, which consists of non-Greek, non-Roma inhabitants of the Balkans:

This appears quite varied; hopefully more Balkan project participants will allow me to split this into additional sample populations.

Finally, here is a portrait of the Ashkenazi population, which appears quite similar to the Behar et al. one:

A very interesting thing about this population is the existence of small slices of "East Asian" and "Northeast Asian" components totalling about 1.5% in almost all individuals. In my opinion this testifies to some type of old minor absorption, as it is fairly evenly spread in the population.

If you haven't joined the project yet, feel free to submit your sample during this opportunity.

Structure in West Asian Indo-European groups

Indo-European languages used to be represented by several branches in West Asia. The most ancient Anatolian languages (Hittite, Luwian, and Palaic) became extinct by ancient times, and its descendants (e.g. Lycian and Lydian) soon thereafter. Greek and Phrygian were added from the Balkans, and so did Armenian, which, according to tradition was derived from the Phrygian.

Unrelated to all these languages, to the east, were the Iranian speakers, the major branches of which extant today in West Asia are Kurdish and Persian (Farsi).

The boundary between the western Indo-Europeans and the east ones changed several times in history. The apex of Iranian power came early when the Achaemenids subjugated the entirety of Asia Minor, but the Iranian expansion was halted in Europe during the Persian Wars of the 5th c. BC. In the next century a reversal of fortunes resulted in the conquest of the Persian Empire by the Greeks of Alexander. The Hellenistic kingdoms lasted for centuries thereafter, but eventually succumbed and new Greco-Persian kingdoms arose that led to the Parthian revival which clashed with the new western power, Rome. The limit between East and West survived with victories and defeats on either side well into medieval times as the eastern Romans fought the Sassanids, the successor dominant power in the Iranian world. Heraclius ended the centuries-old struggle when he defeated the Sassanids in Mesopotamia, but the whole affair became irrelevant as soon thereafter the new power of the Arabs destroyed the Sassanids and threatened the Roman Empire itself which managed to survive for a few more centuries, before eventually succumbing to the Crusaders and Turks, but both of these probably had a minor effect on the local population. Greeks and Armenians continued to exist in Anatolia until the 20th century, with only remnants of them remaining in Asia Minor today.

The data

To study the relationship between the various West Asian Indo-European groups, I gathered an Iranian sample (from Behar et al.), an Iraqi Kurdish one (from Xing et al.), an Armenian one (from Behar et al.), as well as an Armenian one from the Dodecad Project. I have also included the Behar et al. Turkish sample, and a new Turkish sample from the Dodecad Project.

Below are the first two dimensions of the MDS plot.

It appears that Kurds are not particularly closely related to their linguistic cousins, the Iranians. Neither are they very close to Turks and Armenians; the latter appear very close in most of my analyses, with the main difference being a small Mongoloid component in the former, which is not visible here due to the lack of Mongoloid reference populations.

The distinctiveness of the Kurds is also evident in the ADMIXTURE analysis:

The high blue component distinguishes Kurds from both Iranians and Armenians/Turks. Iranians have slightly more of it, suggesting a somewhat closer relationship. The difference between the small Dodecad Turkish sample and the Behar et al. one is suggestive of heterogeneity within Turks, so it is important to be aware of this. Hopefully, if more West Asian individuals join the project, we will be able to discover patterns of regional variation between and within different ethnic groups.

Splitting populations: how many markers needed?

An interesting question in population genetics has to do with the number of markers necessary to infer population-of-origin of an individual.

In the Dodecad Project, we have seen that even very close populations such as Armenians and Assyrians can be classified accurately using the MDS/MCLUST "Clusters Galore" combination that I have proposed. But, do we really need ~177 thousand markers to achieve this level of detail?

I decided to carry out a small experiment, to see how the number of markers used affects the ability to correctly classify samples into two different populations.

Case A: Maximal differentiation (Papuans vs. Mbuti Pygmies)

I begin by considering the case of the two most differentiated human populations in my database, Papuans (17 individuals) and Mbuti Pygmies (13 individuals). For each step I reduce the number of markers by an order of magnitude, using PLINK's --thin 0.1 argument.

• With 176,598 markers, classification is 100% correct, i.e., all 13 Pygmies are assigned to one cluster and all 17 Papuans are assigned to another
• With 17,752 markers, classification is again 100% correct.
• With 1,725 markers, ditto
• With 152 markers, ditto
• With 20 markers, 3 Pygmies are misclassified into the Papuan cluster, hence accuracy is 90%
• With 4 markers, 5 Pygmies are misclassified as Papuans, and 6 Papuans as Pygmies, hence accuracy is 63%

Notice that these are not ancestry informative markers (AIMs) but randomly selected SNPs.

Case B: Small differentiation (Armenians vs. Assyrians)

I have previously shown that a sample of 7 Armenians and 8 Assyrians can be classified correctly, with a single Assyrian misclassified as Armenian, hence 93% accuracy.

• With 17,714 markers, 1 Armenian is misclassified as Assyrian, hence 87%
• With 1,808 markers, accuracy is 80%
• With 188 markers, it is 67%
• With 17 markers, it is 60%

Implications

I carried out this little experiment because I thought it would be interesting, but also for its implications. I can identify at least two major ones:

Ancient DNA: Due to poor preservation we are unlikely to get full genome sequences from ancient human DNA except in highly favorable conditions. In many instances, it may be possible to get only a limited number of markers tested. These results suggest that it is possible to get fairly decent assignment of individuals into populations without hundreds of thousands of SNPs. Thus, it may be possible to study genetic structure in ancient necropoleis, or the relationship of ancient remains to modern populations.

Data integration: Different genotyping platforms (e.g., those of Affymetrix and Illumina) often possess a very small subset of common SNPs. Imputation of genotypes is possible, but these results suggest that even when the overlap between markers is not substantial, it is possible to carry out fairly sophisticated genetic analyses on them.

Conclusion

Naturally, there are issues not addressed here: what happens when we are dealing with more than 2 populations? What happens if we want to study admixture rather than classification? In both cases, I expect the number of markers needed to be higher.

Nonetheless, I am fairly convinced, both from this, and a previous experiment that more markers than are used in current genotyping platforms will only add very little value to anthropological investigations of ethnic genetic differences.

Assyrians vs. Armenians

Assyrians and Armenians have invariably been assigned to the same cluster in Clusters Galore analysis thus far. It is very important to understand the following two axioms:

• Populations that are separated by clustering analysis have enough genetic differences between them to allow such separation
• Populations that are not separated may, or may not have enough genetic differences between them to allow separation

The second axiom is extremely important. Let's elaborate:

• The clustering algorithm may have its limitations
• The number of markers may be insufficient
• The number of individuals may be insufficient

In the Galore approach, I use MCLUST, as good a clustering algorithm as any. I also have samples of 8 Assyrians and 7 Armenians, and use about 177k SNPs, which are enough to squeeze out differences even between related populations. So, what gives? Are Assyrians and Armenians really genetically indistinguishable?

What constitutes a cluster? MCLUST can divide a dataset into as many clusters as you want, but it also chooses the number of clusters to optimize the Bayes Information Criterion. But, it's important to note that the BIC is not some god-given arbiter of what a cluster is; it is best viewed as a guide to choose a good number of clusters, and not as a guarantee that this is the true number of clusters that a population can be subdivided in.

Clustering Assyrians and Armenians, assuming 2 clusters

To that end, I decided to cluster Assyrians and Armenians, forcing MCLUST to infer 2 clusters. I only retained 2 MDS dimensions, as these are enough to distinguish between 2 groups. Here is the MDS plot:

We can observe that Assyrians can be distinguished from Armenians along Dimension 1, and there is also some structure in Assyrians, with 3 of them forming a mini-cluster on top, 4 of them forming another mini-cluster at the bottom, and 1 of them being closer to the Armenians.

By applying MCLUST with K=2, all 7 Armenians and 1 Assyrian are assigned to a cluster, and the remaining 7 Assyrians to another. Thus, the two groups can be separated from each other, although their differences (due to the factors I mentioned) are not large enough to lead to an improvement of the BIC.

With more individuals., it's possible that the BIC too will be able to track the improvement in likelihood that adding two clusters will produce.

And, indeed, it may be possible that some of these populations will be further subdivided. If more Assyrians join the project, for example, it may be the case that the two apparent Assyrian clusters will emerge, or it will be the case that the space between them will be filled by currently unsampled individuals.

As more individuals join the Project, ever-finer distinctions can be uncovered.

NB: I make the case for Assyrians and Armenians, but I have also tried this for other closely related populations who were assigned to the same cluster by Galore analysis, namely Spaniards and Portuguese. In that case, however, I was not able to find a clean solution with K=2. Once again, this may mean that Spaniards and Portuguese are not genetically that different, or it may mean that their difference will be revealed with more individuals joining the project.

How Turkish are Anatolians? revisiting the question

In 2005, I estimated the Y-chromosome heritage of Turkic speakers on modern Anatolians at 11%.

In the same year, I estimated the Mongoloid admixture in Anatolian Turks at 6.2% on the basis of Y-chromosome and mtDNA. This is not inconsistent with the previous percentage, as the Turks, when they arrived in Anatolia were almost certainly of mixed Caucasoid-Mongoloid heritage.

Surprisingly, the maternal contribution from East Eurasia seems higher than the paternal one, on the basis of uniparental markers. But, that is not so surprising if one considers that the Turks who arrived to Anatolia were to a degree descended from Turkicized groups of Iranian steppe nomads bearing Caucasoid patrilineages. Already we have ancient DNA evidence of groups in Central Asia with Caucasoid patrlineages (R1a1) and mixed Caucasoid-Mongoloid mtDNA.

In 2007, some Turkish researchers estimated, using Alu polymorphisms, the Central Asian admixture in Turks at 13%, quite close to my own estimate, and, given the observation that there is more Mongoloid mtDNA than Mongoloid Y-chromosomes in modern Anatolians, the slight difference of 2% is probably taken care of.

In October, I estimated the Mongoloid admixture in Turks at 5.5%, quite close to the 6.2% arrived in 2005 using Y-chromosomes and mtDNA. Subsequently, my K=10 Dodecad analysis (spreadsheet) arrived at 6.7% sum of "East Asian" and "Northeast Asian" components. The slight increase is not surprising, as the K=10 analysis included a greater sampling of Mongoloid diversity.

In ISBA4, another group of Turkish researchers arrived at a 13% estimate for the nomadic Turkic element in modern Anatolian Turks.

Finally, my K=15 analysis has revealed 7.9% "eastern" components in Turks. Given that the "Central Siberian" component is equidistant from Caucasoids and Mongoloids, this translates into about 7.2% East Eurasian admixture. Again, the slightly larger result can be accounted by the sampling of even greater Mongoloid diversity, from the previously unsampled Siberia.

Summary

Y-chromosome, mtDNA, and autosomal DNA analysis by myself and by Turkish researchers all point to 6-7% of Turkish genetic heritage being specifically east Eurasian in origin, and about 1/7 of their genetic heritage coming from Central Asia.

ADMIXTURE analysis

I received an e-mail from a Turkish participant in the project, who wondered whether the K=15 analysis was supportive of much higher demographic influence of Central Asian Turks in the current Turkish population.

In particular, a back-of-the-envelope calculation of "eastern" components in Turks and Uzbeks led him to the conclusion that this was at least 20% and probably more.

Thus, I decided to perform direct ADMIXTURE analysis of Turks and Uzbeks to see what the estimate of Central Asian admixture in Turks actually is.

In the above figure, there are (left-to-right): 1 Dodecad Project 50% Turk-50% Laz showing no Central Asian admixture, 3 Dodecad Project Turks, 19 Turks from Behar et al. (2010), followed by Uzbeks (blue, with some seemingly admixed individuals), followed by 15 Dodecad Project Greeks and Armenians (red).

Behar et al. (2010) Turks have 15.4% Central Asian admixture; if we add the 3 Dodecad Project Turks to the sample, this becomes 14.4%. I'll be happy to tell the three Turks in the Project their individual proportions if they e-mail me.

In conclusion, this analysis too provides an estimate of the Central Asian component in Turks similar to all the ones listed in the beginning.

Conclusion

Estimating the precise genetic identity of nomadic Turks at their time of arrival in Anatolia is difficult to achieve. First of all, modern Anatolian Turks are a subset of recent Anatolians; second, there is the problem of how many Iranian-speakers were absorbed by the westward migrating Turks from Central Asia, and when; also, what was the impact of the Mongol expansion in Central Asia after Turks had already reached the west, and later what were the impacts of Chinese and Russian expansion in the Eurasian heartland.

It's all a big puzzle, but, for the time being 5-7% East Eurasian admixture in modern Anatolian Turks and about 1/7th of their heritage coming from Central Asia seems like a reasonable estimate.

Analysis of Armenians, Lezgins, Georgians, and Adygei

I have taken Behar et al. (2010) Armenians, Georgians, and Lezgins, and HGDP-CEPH Adygei, together with the three Dodecad Project members scoring highest in the "West Asian" component in order to study the structure of Caucasus populations.

The first two dimensions of the MDS plot can be seen on the left. The color coding is as follows: DOD155: green; DOD011: red; DOD049: orange; Adygei: grey; Lezgins: magenta; Georgians: black; Armenians: blue.

It is fairly obvious that Adygei, speakers of a NW Caucasian language, can be distinguished from Lezgins, speakers of a NE Caucasian language, and from Georgians, speakers of a S Caucasian language. Armenians can be distinguished from Georgians, although they appear to be closer to them. The three project members all appear to be in the broad Georgian-Armenian cluster.

Turning to ADMIXTURE, analysis with K=3 confirmed the existence of the afore-mentioned clusters, with Adygei (dark blue), Lezgins (green), and Georgians/Armenians (light blue) forming their own distinctive clusters.

Going one step further at K=4, convergence became more difficult, but, nonetheless, the clusters produced by ADMIXTURE were also informative. Below, you can see the individual-level ADMIXTURE results:

While much "noisier", perhaps due to both the presence of actual admixture, and the difficulty of inferring structure in relatively closely related populations, four clusters emerge: light blue (Adygei), dark green (Lezgins), dark blue (Georgians), and light green (Armenians). The clusters become more apparent if we look at averages:

It is fascinating that four different populations, representative of four different language families (Indo-European and S/NE/NW Caucasian) can be distinguished in the Caucasus. With the exception of Armenians, none of the other populations have linguistic cousins in the world-at-large, although it has been speculated that the extent of Caucasian languages was once much broader than it is today, and some have even suggested that the Neolithic proto-farmers who dispersed into Europe were bearers of such languages.

It will be extremely interesting to study these clusters in other individuals from the Caucasus and Anatolia, as well as to see what is their relationship to the "West Asian" component which appears to be widespread in Eurasia.