Iranians

Since we have 7 Iranians in the project, it's time to look at them as a group. We also have 19 Iranians from the Behar et al dataset.

Let's look at their admixture results at K=12.

The big difference between Harappa Project Iranians and Behar et al Iranians is African admixture. Only one Harappa Iranian (HRP0046) has 1% African admixture while three Behar Iranians have more than 10%.

Let's do hierarchical clustering with complete linkage using the Euclidean distance between admixture components. First a caveat or two. This is not a phylogeny. Also, the Euclidean distance measure is not a good one for measuring differences in admixture but I am not sure what would be better.

HRP0010 who is an Assyrian actually clusters better with Caucasian, Iranian and Iraqi Jews than with Iranians.

I'll run an MDS or PCA of the whole region from Punjab/Kashmir to the Levant and Caucasus soon which should be more interesting for clustering.

UPDATE: Since Palisto wondered, I checked and found out that he, an Iraqi Kurd, is very like the Iranians in his admixture result. So I have included him (HRP0059).

21 Comments.

  1. I am wondering how I (HRP0059) would fit into the Fst divergence diagram since I am also an Iranian, not by nationality though (I am an Iraqi Kurd).

  2. Thank you for creating this.

    It confirms that several of Behar's Iranians have substantial African admixture which is usually absent in Iranians.
    Based on this, it is likely Behar's Iranian sample is actually from either Khuzestan, Balochistan or Hormuzgan/Bandar Abbas (i.e. the southern swathe of the country).

    What I found most surprising is the great variance in European (2-12%) and Southwest Asian (19-43%) components.

  3. Thank you again, Zack. And indeed, DMXX. I reduced Zack's K12 ADMIX components to two dimensions. Although not a PCA or MDS of the genome data per se, the plot is useful in further understanding the relationships between the Harappa and Behar Iranian samples. I have rotated the plot to better represent geography. From among the non-African admixed individuals, I plot farthest west. The Khorasani participant is the most easterly.
    http://tinyurl.com/4e928bc

  4. shorter: iranians = arabs X aryans, pakistanis = brownz X aryans 🙂

    • Are you the actual Razib Khan or a very unconvincing troll?

      The Southwest Asian component (which I presume you're referring to) cannot be equated with "Arabs" when the reference Tuscans, Turks, Spaniards, Sardinians, Georgians and Armenians have values that are in the same ballpark as the Iranian reference average (29%).

      It would be more accurate to describe Iranians as "South Caucasians with an appreciative connection to other Indo-Iranian speakers".

      • take a chill pill. it was partly a joke 😉 though it think to a first approximation the joke is telling us something real.

  5. humayun luwi libi

    I think the southwest asian component is Arabian (i.e from arabian peninsula) and is in most cases a legacy of Arabs (for example in Iran since it's lacking further east in Pakistan that wa not reached by the Arab armies) however in Europe it's most likely Phoenician more than Arab (except Sicily and Spain) and in Africa it's pre Arab various Afrasan speaking populations that did brought afrasan languages to north and east africa, that's my 2 cents.

  6. humayun luwi libi

    " Tuscans, Turks, Spaniards, Sardinians, Georgians and Armenians have values that are in the same ballpark as the Iranian reference average (29%)."

    I think there is some misunderstanding afaik the southwest asian component is very low amongst those population as well as amongst iranians where it's around 8-10% and not 29% perhaps west asian was included too but west asian should not be equated with afrasan speaking arabians but rather with indo-european speaking proto and ancient indo-europeans from anatolia such as mitannis and luwians.

    • west asian should not be equated with afrasan speaking arabians but rather with indo-european speaking proto and ancient indo-europeans from anatolia such as mitannis and luwians

      I thought the mittani were specifically (and surprisingly) Indo-Aryan - as opposed to more generically Indo-European or even Indo-Iranian - at least from a linguistic and religious viewpoint - http://en.wikipedia.org/wiki/Mittani

      • That said, I agree with the broader point. West Asian should not be equated with Arab. Even many of the peoples and territories currently speaking Arabic were quite possibly IE- or Caucasian-speaking earlier.

  7. humayun luwi libi

    By the way statistically speaking Arabs cannot have as much as 29% admixture amongst iranians because they were few and came very lately so at max we may expect 5-10% arabian input.
    To have an idea the Turks that to anatolia came 4 centuries later than arabs and were few numbering military nomads similarly as arabs have around 5% input in anatolia so logically arab input in iran too should not exceed 8-10% and should be a bit higher than the Turkic input in Iran (except for the recent turkmen nomads of northeastern iran)

  8. i guess for the record, my own model here is much deeper in time than what humayun is referring to. so "arabs" means for me a southwest asian neolithic era expansion. "aryans" means a somewhat more recent one. "brownz" themselves are a synthesis of ANI and ASI.

  9. humayun luwi libi

    The neolithic expansion can be traced by west asian component who was brought by the indo-europeans that introduced indo-european languages to iran and india.

    since the southwestern asian component is absent (in behar) amongst pakistanis and indians it could not be neolithic but historical.

    I think you know anatolians have as high as 5% north mongoloid Turk input yet the Turks came very recently to Anatolia (by the 12 th century) so analogously the 7 th century arabs in Iran could be traced by that southwestern asian (of behar and dodecad I dont know about the one here) as 12 th century Turks in Anatolia can be traced by the 5% north mongoloid input.
    south asian component should be dravidic and not indo-european.

    • The ancient mtDNA from Neolithic sites in southern Europe and west Asia showed presence of L(xM,N). South Asia has no L(xM,N) except in populations with recent African and S. West connections, Siddi, Makrani, etc. Which potentially means that the Indo-European input into India was from an area with no L(xM,N).

  10. Hope to see more of these region-specific analyses in the future. 🙂

    [Hint : We have 14 linguistically Tamil people in the project ;)]

  11. to my Irooni baradaran CHILL lol. Razib was meaning on a deeper timescale and its always been established that Iranians are a Mid-East population with a thin "Aryan" veneer, thinner anyway west of Dasht-e-Lut and Dasht-e-Kavir. Its also pretty obvious if you "look" at Iranians, North Iranians & Azeris are fair like Caucasians (from the Caucasus), whereas South Iranians are swarthier and Iranians from Khorasan sort of sometimes have an Afghanish look.

    However when the figures spike up beyond the ref range of mid 20's-low 30's its probably down to the Bedouin contribution.

    Iranians are always going to be mixed because it was the default assimilatory ethnicity that all migrants/conquerors of the Iranian plateau adopted. Interestingly enough this is only the Persians, I'd be interested in the Azeris (can we make Mid-Easterners technically brown for the purposes of this project), Zoroastrians, Qashqais and other ethnicities to sort of supplement Iranian data.

    Also as the great Palisto comment (comment #1) "I am also an Iranian". Iranianism completely transcends genes so the descendant of a Bedouin or African or Turk, who feels the passion for Iran & Iranism, is as Iranian as the descendant of a Zoroastrian.

    @ Vasishta I completely agree we need further context rather than only linguistic/national. I mean its been brilliant so far credit to Zack (what is it about Zacs??)..

  12. humayun luwi libi

    "The ancient mtDNA from Neolithic sites in southern Europe and west Asia showed presence of L(xM,N). South Asia has no L(xM,N) except in populations with recent African and S. West connections, Siddi, Makrani, etc. Which potentially means that the Indo-European input into India was from an area with no L(xM,N)."

    Genetic drift, it's easy to assume a single tribe from a particular region (with no L mt DNA expanding into india).

  13. anatolians have as high as 5% north mongoloid

    You are doing the same mistake that I warned about in my first posts on this blog. The reference populations are very small and usually come from obscure locales, so they have a problem of representativeness; this is especially true when it comes to minority components (like the one you mention). In the case of Turks, their reference population consists of only 19 individuals from Behar et al. and their locale(s) is/are completely obscure (they may even be from a single locale of Turkey, as they are rather genetically homogeneous compared to the Dodecad, Eurogenes and 23andMe Turkish samples).