Total Pageviews

Wednesday, April 17, 2013

Kurdish mtDNA data IX

Just an update (N=97):


1x B4b1a (Kurd from Turkey) 
1x C4b (=C4b+A248G, A14566G, T16519C, A249A (not deleted)) (Alevi Kurmanji)
1x G2a (=G2a+T16172C) (Sorani)
1x H CRS (Kurds from Iran; Quintana-Murci et al., 2004)
6x H (Kurds from Georgia; Comas et al., 2000)
2x H with T16311C  (Kurds from Georgia; Comas et al., 2000)
1x H with C16218T (=H1ag1a or H1aq1 or H20) (Kurds from Georgia; Comas et al., 2000)
1x H with C16192R, C16261T (similar to: JN415470(Italy-LHON) Achilli Haplogroup H 19-AUG-2012 (Kurds from Georgia; Comas et al., 2000)
1x H5'36 (Kurd from Turkey)
1x H5  16051G, 16255A, 16304C, 16319A, 16327T, 263G, 315.1C, 456T (Kurds from Iraq; Al-Zahery et al., 2012)   
1x H5a1 (H5a1+T16304G, A3397G, G5471A) (G5471A usually in H5b) (Sorani)
1x H13a2b2 (Alevi Kurmanji from Dersim)
1x H14b T3197C (=H14b+C4086T, A16265T) (Yezidi)
1x H14a with T16311C, C16256T, T16352C (=H14a +T16311C ) (Kurds from Georgia; Comas et al., 2000)
1x H15a1 (=H15a1+309.1C, 309.2C, 315.1C, A15316G) (Sorani; mtDNA fully sequenced here and here)
1x H15b (Sorani)
1x H15b T16086C (close to EU600353(Druze) Shlush)(Kurds from Iran; Quintana-Murci et al., 2004)
1x HV* CRS (Kurds from Iran; Quintana-Murci et al., 2004)
1x HV* 16174 (Kurds from Iran; Quintana-Murci et al., 2004)
1x HV1a1 C16067T, C16355T (=HV1a1) (Kurds from Iran; Quintana-Murci et al., 2004)
1x HV2 C16168T, T16189C, T16217C, C16287T (Kurds from Iran; Quintana-Murci et al., 2004)
1x HV14 T16311C, G4655A, T15115C (Sorani)
1x HV A73G, T391C, G16153A (Kurmanji from Zakho)
1x HV (Kurmanji from Diyarbakir (Amed)/Turkey)
1x I1a G16129A, C16168T, T16172C, 16173, C16223T (Kurds from Iran; Quintana-Murci et al., 2004)
1x I with G16129A, C16223T (Kurds from Georgia; Comas et al., 2000)
1x I1a1d with G16129A, C16223T, T16172C, T16189C, C16083T, C16355T (=I1a1d+C16083T, C16355T) (Kurds from Georgia; Comas et al., 2000)
1x I5a pre-I5a3  because G5231A, A15052G = I5a3 but still C150C, T6278T = I5a; additional   C16301T (Zaza from Dersim)
1x I5a (Zaza from Baltas/Varto, Turkey)
1x J1b C16069T, T16126C, G16145A, C16222T, A16235G, C16261T, T16311C (Kurds from Iran; Quintana-Murci et al., 2004)
1x J1b3 T1822C, A8460G, T16311C  (=J1b3+T1822C+T16311C) (Sorani)
2x J1b1b1 C16069T, T16126C, G16145A, C16261T, 16519C, A73G, A263G, C295T, 309.1C    315.1C, C462T, T489C, 523DEL, 524DEL   (very close to JF939049(Armenian)) (Kurds from Iraq; Al-Zahery et al., 2012)
1x J1b3b A73G, A263G, C295T, T489C, A750G, A1438G, A2706G, G3010A, T4216C, A4769G, C7028T, G8269A, A8460G, A8860G, A10398G, A11251G, G11719A, A12612G, G13708A, C14766T, A15326G, C15452A, T15530C, C16069T, T16126C, G16145A, C16222T, A16235G, C16261T (1/2 Alevi Kurmanji paternally, 1/2 Sunni Kurmanji maternally from Bingol, Kighi, Turkey)
1x J1b4 C16069T, T16126C, G16145A, C16222T, C16261T C16278T, C16287T (Kurds from Iran; Quintana-Murci et al., 2004)
1x J1c (=J1c+G185T, 4812A, C16290T, T16519C) (Alevi Kurmanji from Dersim)
1x J1c2m (old J1c2a)  C16069T, T16126C, 16148T, A73G, 185A, 228A, A263G, C295T, 315.1C, C462T, T489C, 523DEL, 524DEL  (close to JQ797801 from Romania and  JQ797802 from West-Siberia (Khanty))(Kurds from Iraq; Al-Zahery et al., 2012)  
1x J1d (=J1d+A15218G, T16519C) (Feyli, originally from Iran)
1x J1d (Kurd from Iraq)
1x J2a1a1 (=J2a1a1+A10044G, G11914A, C16264T)  (Kurd from Turkey)
1x J2b1 (Kurd from Iraq/Iran)
1x JT with T16126C, C16067T, T16311C (=JT) (Kurds from Georgia; Comas et al., 2000)
1x K1a T16093C, T16224C, T16311C (Kurds from Iran; Quintana-Murci et al., 2004)
1x K 16129, T16224C, T16311C (= K1a11 or K2b2) (Kurds from Iran; Quintana-Murci et al., 2004)
1x K with T16224C, T16311C (Kurds from Georgia; Comas et al., 2000)
2x K with T16224C, T16311C, T16093C, C16260T (=K1a1+C16260T, or K1a17a+T16093C)(Kurds from Georgia; Comas et al., 2000)
1x K with T16224C, T16311C, A16240G (=K+A16240G) (Kurds from Georgia; Comas et al., 2000)
1x K with T16224C, T16311C, A16272G (=K+A16272G) (Kurds from Georgia; Comas et al., 2000)
1x L3e5 16037G, A16041G, C16223T, A73G, C150T, A263G, 315.1C, T398C, 523DEL, 524DEL
(Kurds from Iraq; Al-Zahery et al., 2012)
1x M1a1 G16129A, 16182C, 16183C, T16189C, C16223T, T16249C, T16311C, T16359C, C16360T, T16519C, A73G, T195C, A263G, 309.1C, 309.2C, 315.1C, T489C (Kurds from Iraq; Al-Zahery et al., 2012)
1x N1b1 (=N1b1+C16176C, C1703A, C3921A, G7337A, T16519C) (Alevi Kurmanji from Dersim)
1x N1b1 with C16223T, G16145A, C16176G (=N1b1) (Kurds from Georgia; Comas et al., 2000)
1x N2a with C16223T, T16086C, G16153A, G16319A (=N2a+T16086C) (Kurds from Georgia; Comas et al., 2000)
1x R0 16519C, 16524C, A263G, 315.1C (Kurds from Iraq; Al-Zahery et al., 2012)
1x R0 16368C, 16519C, A263G, 309.1C, 315.1C (Kurds from Iraq; Al-Zahery et al., 2012)   
1x R2 with C16071T, G16145A, C16234 (=R2 +G16145A+C16234) (Kurds from Georgia; Comas et al., 2000)
1x T1 (=T1a2b+C12633T, G5460A, G11914A, T16311C, (T16519C))(Feyli)
1x T1a7 T16126C, A16163G, C16186T, T16189C, G16274A, C16294T, T16519C, A73G, A263G, 309.1C, 315.1C, A512G  (close to EU935435(Egypt) Kujanova and JQ798027 (Israel))(Kurds from Iraq; Al-Zahery et al., 2012)
1x T1b T16126C, A16163G, T16189C, T16243C, A16247G, C16294T, T16519C, A73G, 152C, A263G, 309.1C, 315.1C, 524.1A, 524.2C (Kurds from Iraq; Al-Zahery et al., 2012)
1x T2a1b2b with T16126C, C16294T, C16296T, C16256T, A16317G (=T2a1b2b +A16317G) (Kurds from Georgia; Comas et al., 2000)
1x T2b (Kurd from Turkey)
1x U1a1 A14070A, T16163C (Zaza)
1x U1a1a (Sorani) with A11467G,  A12308G,  G12372A, C285T,  T12879C,  A13104G, A14070G, G15148A, A15954C, T16249C, C2218T, G14364A, T16189C, G4991A, G6026A, T7581C, A385G, 3158.1T, G3591A, A13422G, G9575A, C2836T, G4659A, 573.1C, 573.2C, A10283G, (309.1C), (315.1C), (523-), (524-), (16182C), (16183C), (16519C) (=U1a1a; shares C2836T, G4659A mutation with Indian samples HM156682(India) Govindaraj) (fully sequenced here and here)
1x U1a1 with A16182C, A16183C, T16189C, T16249C (=U1a1) (Kurds from Georgia; Comas et al., 2000)
1x U1b C16111T, 16214A, T16249C, G16319A, C16327T, T16519C, A73G, T146C, T152C, A263G, C285T, 315.1C, 572T (Kurds from Iraq; Al-Zahery et al., 2012)
1x U2 (Alevi with Zaza ancestry)
1x U2e1a with A16051G, T152C, A508G, A3720G, A5390G, T5426C, C6045T, T6152C, A10876, T13020C, T13734C, A15907G, G16129C, T16362C, C340T, C11197T, T11732C, G7337A, A15218G, T16311C, T16519C (Sorani from Sulaymaniyah/Iraq)
1x U3 with A16343G (=U3) (Kurds from Georgia; Comas et al., 2000)
1x U3a with C150T, A14139G,  T15454C, A2294G,  T4703C,  G9266A, T6518A (flip!),  A10506G,  C13934T,  G16390A C2766A, 10790C, G16129A, 16257T, T16519C (Sorani from Sulaymaniyah/Iraq); (similar to HM852895(Georgian45) Schoenberg also with C2766A)
1x U3c with A16343G, C16193T, T16249C (=U3c) (Kurds from Georgia; Comas et al., 2000)
1x U3c C16193T, T16249C, A16343G, G16526A, A73G, C150T, A263G, 315.1C  (close to HM852797(Azeri34) and HM852803(Azeri42) Schoenberg)(Kurds from Iraq; Al-Zahery et al., 2012)   
1x U4 T16356C, T16519C, A73G, T195C, A263G, 309.1C, 315.1C, G499A, 524.1A, 524.2C
(Kurds from Iraq; Al-Zahery et al., 2012)
1x U5 T16093C, T16189C, C16270T (Kurds from Iran; Quintana-Murci et al., 2004)
1x U5a1 (Kurmanji from Dohuk)
1x U7 (Zaza from Turkey)
1x U7 C16069T, A16227G, C16278T, A16318C, T16359C (very close to HM852853(Turk 187) Schoenberg) (Kurds from Iran; Quintana-Murci et al., 2004)
1x U7 C16192T, A16309G, A16318T (Kurds from Iran; Quintana-Murci et al., 2004)
1x U7 T16243C, A16309G, A16318T (Kurds from Iran; Quintana-Murci et al., 2004)
1x U7 A16309G, A16318T (Kurds from Iran; Quintana-Murci et al., 2004)
1x U7 A16309G, A16318T (Kurds from Georgia; Comas et al., 2000)
1x U7a A16309G, A16318T, T16519C, A73G, C151T, T152C, A263G, 309.1C, 315.1C, 523DEL, 524DEL (Kurds from Iraq; Al-Zahery et al., 2012)
1x U8b1a1 A16066G, G16129A, C16169T, A16183C, T16189C, C16234T, T16311C (Kurds from Iran; Quintana-Murci et al., 2004)
1x U8b C16111T, T16172C, A16183C, T16189C, T16311C (Kurds from Iran; Quintana-Murci et al., 2004)
1x U8b (Feyli)
1x U8b (Zaza from Sivas, originally from Dersim) 
1x W C16223T, C16292T (Kurds from Iran; Quintana-Murci et al., 2004)
1x W3 G16153A, C16223T, C16292T, C16294T, T16519C, A73G, T152C, A189G, C194T, T195C, T199C, T204C, G207A, A263G, 309.1C, 315.1C (Kurds from Iraq; Al-Zahery et al., 2012)
1x W4a C16223T, C16292T, C16286T (Kurds from Iran; Quintana-Murci et al., 2004)
1x W6 with C16292T, C16192T, C16223T,  T16324C (=W6) (Kurds from Georgia; Comas et al., 2000)
1x X with T16189C, C16278T, C16186T (=X+C16186T) (Kurds from Georgia; Comas et al., 2000)

Related post:
mtDNA of Kurds Part I
mtDNA of Kurds Part II
mtDNA of Kurds Part III
mtDNA of Kurds Part IV
mtDNA of Kurds V
Kurdish mtDNA data VI 
Kurdish mtDNA data VII
Kurdish mtDNA data VIII

36 comments:

  1. Hello Palisto,

    In light of all the autosomal, Y-chromosomal and mitochondrial analyses involving Kurds so far, are the main branches of Kurds genetically indistinguishable from each other or do they have peculiarities?

    ReplyDelete
    Replies
    1. Looking at autosomal data, the different branches form a continuum, a nice gradient: Zaza>Kurmanji>Sorani>Gorani-Feyli.

      The Northern samples have more of the components Caucasus, Atlantic-Mediterranean/Mediterranean and N-European/NE-European, while the Southern samples have more of the 2 components Gedrosia/Baloch and SW-Asian.

      Looking at Y-Chromosome haplogroups, I see some peculiarities but this could be due the low N per region. Anyways, the differences between Kurds and Zazas from Iran, Iraq and Turkey.

      Iran: J2 > E > J1 > R1a > G
      Iraq: J2 > R1a > J1 > E/R1b
      Turkey: J2 > R1a/J1/I > E

      Haplogroup J2 dominates in Kurds;
      haplogroup R1b appears mostly in Iraqi Kurds;
      haplogroup I appears mostly in Zaza/Kurmanjis from Turkey;
      haplogroup G and E appears to be more in Iranian Kurds.
      One Kurdish Y-haplogroup G individual from Turkey has oral stories about coming from Iran, which fits to this observation.

      Regarding mtDNA, the picture is even less clear. Iranian Kurds tend to have more U7, U8, K, J1, HV and W. Iraqi Kurds tend to have more J1, H5, R0, T1, U1, and U3. H15 seems to be present in the Iranian/Iraqi border region.

      Delete
    2. Thanks for the detailed info. In Y-chromosomes what differentiates both Kurds and Iranians from all other West Asian populations is that they have more R1a than R1b, which is the opposite of the West Asian norm. In autosomal DNA what differentiates both Kurds and Iranians from all other West Asian populations is their elevated South Asian component(s) relative to the West Asian standards.

      Do the main Kurdish branches form a Kurdish-specific genetic cluster relative to all other West Asian populations (including the Iranic ones), or, is the Kurdishness more a matter of culture rather than genetics, especially vis-à-vis the non-Kurdish Iranic West Asians? If the latter is the case, that supports the theory that the Kurdish ethnic identity gradually formed during the Middle Ages from the interactions (with each other and the outside world) of the various dispersed Iranic-speaking tribes of northwestern Persia/Iran with a similar lifestyle despite the (still) low level of mutual intelligibility between the Iranic language of some of the participant tribes.

      Delete
    3. "In Y-chromosomes what differentiates both Kurds and Iranians from all other West Asian populations is that they have more R1a than R1b, which is the opposite of the West Asian norm."

      I am not exactly sure what your definition of "West Asian populations" and "West Asian norm" is.
      All Muslim populations East of Euphrates and all Muslim populations of the Arabian peninsula have more R1a than R1b.

      Turks as well as the Jewish and Christian minorities have more R1b than R1a.

      "In autosomal DNA what differentiates both Kurds and Iranians from all other West Asian populations is their elevated South Asian component(s) relative to the West Asian standards. "

      In Dodecad K12B Kurds_D and Iranians_D have a South Asian component of 1.1% and 3.6%, respectively. Syrians have 1.3%, Yemenese have 2.4%, Turks_D have 1% of the South Asian component. This is not a remarkably important component because the differences to others are marginal.

      Clearly, the most important component that distinguishes Kurds and Iranians from other West Asian populations is the Gedrosia component; Kurds_D and Iranians_D have 28.7% and 28.8%, respectively, while all their Western neighbors have below 19%.

      "Do the main Kurdish branches form a Kurdish-specific genetic cluster relative to all other West Asian populations (including the Iranic ones), or, is the Kurdishness more a matter of culture rather than genetics, especially vis-à-vis the non-Kurdish Iranic West Asians?"

      Dienekes observed a Kurdish-specific cluster.
      http://1.bp.blogspot.com/_Ish7688voT0/TR8ox_MI6qI/AAAAAAAADIE/zEcyBpR0U8s/s1600/MDS1600.png
      In the image above Zaza people would end up in the top left area of the Kurdish cluster (close to Caucasus cluster), while Soranis would end up in the bottom right area of the Kurdish cluster (close to Iranian cluster).

      Delete
  2. Just another map made by Dienekes that shows a Kurdish cluster
    http://3.bp.blogspot.com/-e5L7sf6mEcY/TcCBaZ9rFqI/AAAAAAAADnM/Rd7jcR8XKYI/s1600/1_2.png

    From:
    http://dienekes.blogspot.com/2011/05/solution-to-problem-of-indo-aryan.html

    ReplyDelete
  3. I am not exactly sure what your definition of "West Asian populations" and "West Asian norm" is.
    All Muslim populations East of Euphrates and all Muslim populations of the Arabian peninsula have more R1a than R1b.

    Turks as well as the Jewish and Christian minorities have more R1b than R1a.


    Sorry, I should have said "except the populations of the Arabian Peninsula". Your east of the Euphrates claim must be wrong, as Iraqi Arabs have more R1b than R1a according to the extant genetic studies.

    In Dodecad K12B Kurds_D and Iranians_D have a South Asian component of 1.1% and 3.6%, respectively. Syrians have 1.3%, Yemenese have 2.4%, Turks_D have 1% of the South Asian component. This is not a remarkably important component because the differences to others are marginal.

    Clearly, the most important component that distinguishes Kurds and Iranians from other West Asian populations is the Gedrosia component; Kurds_D and Iranians_D have 28.7% and 28.8%, respectively, while all their Western neighbors have below 19%.


    That is because the "Gedrosia" component of Dienekes captures part of the South Asian genes, clearly more so than the "West Asian" and "Caucasus" components. See:

    http://3.bp.blogspot.com/-o3aBKiyqagc/TyfglQUIlPI/AAAAAAAAEdI/LQj6qi7LE8k/s1600/1_2.png

    ... and see the component results of the West Asian populations at:

    https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedGR2ZWRoQ0VaWTc0dlV1cHh4ZUNJRUE&pli=1#gid=17

    http://www.dnatribes.com/dnatribes-snp-admixture-2013-02-11.pdf

    ReplyDelete
    Replies
    1. "Your east of the Euphrates claim must be wrong, as Iraqi Arabs have more R1b than R1a according to the extant genetic studies."

      Let me explain my thought process:
      First, I specifically wrote "Muslim populations East of Euphrates".

      I don't know of any study that specifically analyzed Iraqi Arabs, only Iraqi people in general. Al-Zahery et al. analyzed Iraqis that had 10.8% R1b and 6.5% R1a1a. However, the used samples had 13.7% Assyrians (Assyrians are not Arabs; Assyrians are not Muslims but Christians). From other sources we know that approximately 25% of all Assyrians are R1b.
      http://www.anthrogenica.com/showthread.php?615-Assyrian-Y-DNA-Distribution/page9
      So, in Al-Zahery et al. approximately 3.4% (=13.7% x 25%) of the R1b in the Iraqi samples are actually derived from the (over-represented) Christian Assyrians. In other words the percentage of R1a and R1b in Muslim Iraqis is 7.4% R1b and 6.5% R1a, this is nearly the same percentage. Considering that the river Euphrates runs through the middle of Iraq and considering that R1b frequencies go up in Western neighbors of Iraq and R1a frequencies go up in Eastern neighbors of Iraq, it is just fair to assume that "All Muslim populations East of Euphrates have more R1a than R1b."

      "That is because the "Gedrosia" component of Dienekes captures part of the South Asian genes, clearly more so than the "West Asian" and "Caucasus" components. See:

      http://3.bp.blogspot.com/-o3aBKiyqagc/TyfglQUIlPI/AAAAAAAAEdI/LQj6qi7LE8k/s1600/1_2.png

      ... and see the component results of the West Asian populations at:

      https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedGR2ZWRoQ0VaWTc0dlV1cHh4ZUNJRUE&pli=1#gid=17

      http://www.dnatribes.com/dnatribes-snp-admixture-2013-02-11.pdf"

      I could simply argue that by saying: "That is because the "South Asian" component of DNA tribes captures part of the "Gedrosia" genes."

      Depending on the number of components in an ADMIXTURE analysis and depending on who is defining these components (and reference populations) the "epicenter" of these components can shift more East or West. Clearly, Dienekes has way more reference populations, including multiple Kurdish sample sets, so his ADMIXTURE runs are more reliable, especially for the Middle East.
      A few months ago, the same "DNA tribes" company presented an analysis where the "Persian" component was peaking in Kurds, while the "Indus Valley" component was higher in the Caucasus people than in Kurds and the "South India" component was 0.0% for Kurds...
      http://kurdishdna.blogspot.com/2012/07/dna-tribes-presents-kurdish-data.html

      Delete
    2. Let me explain my thought process:
      First, I specifically wrote "Muslim populations East of Euphrates".

      I don't know of any study that specifically analyzed Iraqi Arabs, only Iraqi people in general. Al-Zahery et al. analyzed Iraqis that had 10.8% R1b and 6.5% R1a1a. However, the used samples had 13.7% Assyrians (Assyrians are not Arabs; Assyrians are not Muslims but Christians). From other sources we know that approximately 25% of all Assyrians are R1b.
      http://www.anthrogenica.com/showthread.php?615-Assyrian-Y-DNA-Distribution/page9
      So, in Al-Zahery et al. approximately 3.4% (=13.7% x 25%) of the R1b in the Iraqi samples are actually derived from the (over-represented) Christian Assyrians. In other words the percentage of R1a and R1b in Muslim Iraqis is 7.4% R1b and 6.5% R1a, this is nearly the same percentage. Considering that the river Euphrates runs through the middle of Iraq and considering that R1b frequencies go up in Western neighbors of Iraq and R1a frequencies go up in Eastern neighbors of Iraq, it is just fair to assume that "All Muslim populations East of Euphrates have more R1a than R1b."


      We don't know the regional genetic variation in Iraqi Arabs, so the matter is open to some speculation, BUT we do know from the extant genetic studies that both Marsh Arabs (an Iraqi Arab group from southeastern Iraq) and Iranian Arabs have more R1b than R1a, so this strongly suggests that Iraqi Arabs east of the Euphrates have more R1b than R1a.

      I could simply argue that by saying: "That is because the "South Asian" component of DNA tribes captures part of the "Gedrosia" genes."

      Depending on the number of components in an ADMIXTURE analysis and depending on who is defining these components (and reference populations) the "epicenter" of these components can shift more East or West. Clearly, Dienekes has way more reference populations, including multiple Kurdish sample sets, so his ADMIXTURE runs are more reliable, especially for the Middle East.
      A few months ago, the same "DNA tribes" company presented an analysis where the "Persian" component was peaking in Kurds, while the "Indus Valley" component was higher in the Caucasus people than in Kurds and the "South India" component was 0.0% for Kurds...
      http://kurdishdna.blogspot.com/2012/07/dna-tribes-presents-kurdish-data.html


      DNA Tribes has a very rich database, so does Dienekes. There is nothing in the analyses of both of them that contradicts what I have been saying.

      As for the DNA Tribes analysis you linked to, apparently the "Persian" and especially the "Balochi" components capture some of the South Asian genes.

      Delete
    3. "We don't know the regional genetic variation in Iraqi Arabs, so the matter is open to some speculation, BUT we do know from the extant genetic studies that both Marsh Arabs (an Iraqi Arab group from southeastern Iraq) and Iranian Arabs have more R1b than R1a, so this strongly suggests that Iraqi Arabs east of the Euphrates have more R1b than R1a."

      Not sure where you got that information about Iranian Arabs but Iranian Arabs have way more R1a than R1b. See Nasidze et al., 2008:
      http://www.ncbi.nlm.nih.gov/pubmed/18205892
      R-M17 (R1a1a)= 11%
      R-M173 (R excluding R1a1a)=4%

      The Marsh Arabs are a very small minority of Muslim Iraqis (125,000-150,000 in Iraq based on Wikipedia), they don't represent Muslim Iraqis but themselves. Due to their small population size they show a dramatically reduced haplogroup diversity. Thus, their Y-haplogroup diversity don't matter at all, only their STR values are sort of interesting.

      In this study Marsh Arabs also lack J diversity (0% J2-M47, 0% J2-M410, 0% J2-M67, 0% J2-M92, 0% J2-M68, 0% J2-M12, J2-M241) but they have 73% (!!!) J1-PAGE08. Of course they also lack R diversity compared to their neighbors.

      Al-Zahery et al., 2011 (the famous study about Marsh Arabs) wrote about the Marsh Arab:
      "A higher male than female homogeneity is characteristic of the Marsh Arab gene pool, LIKELY DUE TO A STRONG MALE GENETIC DRIFT determined by socio-cultural factors (patrilocality, polygamy, unequal male and female migration rates). "


      "As for the DNA Tribes analysis you linked to, apparently the "Persian" and especially the "Balochi" components capture some of the South Asian genes."

      Let's agree to disagree here.

      Delete
    4. Not sure where you got that information about Iranian Arabs but Iranian Arabs have way more R1a than R1b. See Nasidze et al., 2008:
      http://www.ncbi.nlm.nih.gov/pubmed/18205892
      R-M17 (R1a1a)= 11%
      R-M173 (R excluding R1a1a)=4%


      My source is Grugni et al. 2012, which shows more R1b than R1a in Iranian Arabs.

      Let's agree to disagree here.

      It is only natural for the Iranic peoples to carry more South Asian admixture than their non-Iranic neighbors to the west and north, as Iranic peoples are geographically neighbors of South Asians and linguistically close relatives of northern South Asians. Here is the ADMIXTURE result of another study (Yunusbayev et al. 2011) that, as usual, shows more South Asian genetic element in Iranians and Kurds than non-Iranic West Asians:

      http://2.bp.blogspot.com/-AiqxfdH1kqo/TnC1DOX7yUI/AAAAAAAAEHk/LanedZAS19Q/s1600/admixture-caucasus.png

      Of course, Iranians and Kurds are genetically West Asian peoples, but that does not mean they have no peculiarity and outside genetic influence.

      Delete
    5. "My source is Grugni et al. 2012, which shows more R1b than R1a in Iranian Arabs."

      I see. All these studies with N<100 for each population should be taken with a grain of salt. In this Grugni et al. study Persians, Gilakis, Lurs, and Turkmen have more R1b than R1a as well.
      http://3.bp.blogspot.com/-9ItFg3ZDOCc/UAfrJKKY7aI/AAAAAAAAFC8/WQfxUO6_9Vw/s1600/journal.pone.0041252.t001.jpg


      "It is only natural for the Iranic peoples to carry more South Asian admixture than their non-Iranic neighbors to the west and north, as Iranic peoples are geographically neighbors of South Asians and linguistically close relatives of northern South Asians. "

      It's only natural for the Iranic peoples to share partially some common genetic ancestry with South Asians but that does mean that this common ancestry is South Asian. Your argumentation kinda implies that Iranian languages/people migrated from South Asia to West Asia, which is not the case.

      How about these explanations:
      1. Neolithic revolution with R1b-lacking people in the Fertile Crescent.
      2. Expansion of R1b-free farmers from the Fertile Crescent to North, East, South, and West. At one uncertain time, Neolithic migration to the West (Anatolia/Europe) absorbed R1b people. Contrary, Neolithic migration to the East (South Asia) did not absorb R1b people at any time.

      And/Or

      1. Indo-Iranian languages/people started in Central Asia.
      2. Migration into Iran and India and split into Iranian and Indo-Aryan languages/people due to geographical isolation.
      3. Genetic input of extinct Indo-Iranian people visible in Iranians and South Asians.

      And/Or

      1. Indo-Iranian languages/people started in Iran.
      2. Migration from Iran into South Asia and split into Iranian and Indo-Aryan languages/people due to geographical isolation.
      3. Genetic input of extinct Indo-Iranian people visible in Iranians and South Asians.

      And/Or

      1. Elam-Dravidian languages/people started in Iran.
      2. Migration from Iran to South Asia and split into Elam and Dravidian languages/people due to geographical isolation.
      3. Genetic input of extinct Elam-Dravidian people visible in Iranians and South Asians.

      Delete
    6. Note that in DNAtribes Indus Valley component reaches in Dargins and Lezgians as high as among Kurds

      Delete
  4. All these studies with N<100 for each population should be taken with a grain of salt.

    The Nasidze et al. 2008 study you use as a proof of more R1a than R1b in Iranian Arabs studied 46 Iranian Arabs and all of them were sampled from a single city of Iran. The Grugni et al. 2012 study I use as a proof of more R1b than R1a in Iranian Arabs studied 57 Iranian Arabs and they were sampled from different parts of the Khuzestan Province of Iran. So the Grugni et al. Iranian Arabs are more representative of Iranian Arabs than the Nasidze et al. Iranian Arabs are.

    In this Grugni et al. study Persians, Gilakis, Lurs, and Turkmen have more R1b than R1a as well.
    http://3.bp.blogspot.com/-9ItFg3ZDOCc/UAfrJKKY7aI/AAAAAAAAFC8/WQfxUO6_9Vw/s1600/journal.pone.0041252.t001.jpg


    Only the Persian sample of the Fars Province and the Zoroastrian Persian sample of Tehran have more R1b than R1a, all other Persian samples of the Grugni et al. 2012 study have more R1a than R1b, so it indicates that Persians as a whole have more R1a than R1b. The Grugni et al. Turkmens have much more R1a than R1b. The Gilaki and Lur peoples are from western Iran, so the presence of more R1b than R1a in them and two Persian samples from western Iran is indicative of the R1b-R1a line (roughly speaking) passing through western Iran rather than Iraq, contrary to your claim.

    It's only natural for the Iranic peoples to share partially some common genetic ancestry with South Asians but that does mean that this common ancestry is South Asian. Your argumentation kinda implies that Iranian languages/people migrated from South Asia to West Asia, which is not the case.

    How about these explanations:

    .
    .
    .


    Iranians and Kurds are genetically closer to South Asians than all other West Asian peoples are, and by South Asians I mean all South Asians regardless of language and region, not just the Indo-Iranian-speaking ones or the Indo-Iranian-speaking + Dravidian-speaking ones. There is surely an ASI-related connection. As in Iranians and Kurds, the ASI-related South Asian genes are found in more than noise proportions in Central Asia, the putative homeland of Indo-Iranians.

    ReplyDelete
    Replies
    1. "The Nasidze et al. 2008 study you use as a proof of more R1a than R1b in Iranian Arabs studied 46 Iranian Arabs and all of them were sampled from a single city of Iran. The Grugni et al. 2012 study I use as a proof of more R1b than R1a in Iranian Arabs studied 57 Iranian Arabs and they were sampled from different parts of the Khuzestan Province of Iran. So the Grugni et al. Iranian Arabs are more representative of Iranian Arabs than the Nasidze et al. Iranian Arabs are."

      The number of tested individuals in both studies are too small to make bold assumptions. Let's take the sum:

      Grugni et al.: 57 Iranian Arabs
      3 individuals with R1b, 2 individuals with R1a

      Nasidze et al.: 46 Iranian Arabs
      2 individuals with R1b (or R2!), 5 individuals with R1a

      Sum: 103 Iranian Arabs
      Max. 5 individuals with R1b, 7 individuals with R1a
      R1b: max. 5%
      R1a: 7%

      "Only the Persian sample of the Fars Province and the Zoroastrian Persian sample of Tehran have more R1b than R1a, all other Persian samples of the Grugni et al. 2012 study have more R1a than R1b, so it indicates that Persians as a whole have more R1a than R1b."

      The number of tested individuals in each province are too small, so let's take the sum:

      Grugni et al.: 44 Persian from Fars
      5 individuals with R1b, 2 individuals with R1a

      Grugni et al.: 11 Persian from Isfahan
      0 individuals with R1b, 2 individuals with R1a

      Grugni et al.: 59 Persian from Khorasan
      2 individuals with R1b, 13 individuals with R1a

      Grugni et al.: 46 Persian from Yazd
      1 individuals with R1b, 6 individuals with R1a

      Sum: 160 Persians
      8 individuals with R1b, 23 individuals with R1a
      R1b: 5%
      R1a: 14%

      Grugni et al.: 47 Zoroastrians from Yazd and Tehran
      4 individuals with R1b, 6 individuals with R1a
      R1b: 9%
      R1a: 13%

      "The Gilaki and Lur peoples are from western Iran, so the presence of more R1b than R1a in them and two Persian samples from western Iran is indicative of the R1b-R1a line (roughly speaking) passing through western Iran rather than Iraq, contrary to your claim. "

      The Gilaki people are from the Gilan province. The Gilan province is not in Western Iran as you stated but Northern Central Iran, not far away from the capital Tehran. Furthermore, you stated that Iranians [i]"have more R1a than R1b, which is the opposite of the West Asian norm"[/i]. Well, Lur and Gilaki are obviously Iranians and show less R1a than R1b, so either your statement at the beginning of our conversation is incorrect and/or all of these studies with N<100 for each population should be taken with a grain of salt. I prefer the later option.

      I wrote that R1b/R1a difference is mostly along the different religious lines with Muslims of West Asia having more R1a than R1b, and Jews and (former) Christians having more R1b than R1a.

      The Christian minorities of Iran:

      Grugni et al.: 48 Christian Assyrians from West Azerbaijan and Tehran
      14 individuals with R1b, 4 individuals with R1a

      Grugni et al.: 34 Christian Armenians from Tehran
      8 individuals with R1b, 1 individuals with R1a

      Sum of Christians: 82 Christians
      22 individuals with R1b, 5 individuals with R1a
      R1b: 27%
      R1a: 6%

      "Iranians and Kurds are genetically closer to South Asians than all other West Asian peoples are, and by South Asians I mean all South Asians regardless of language and region, not just the Indo-Iranian-speaking ones or the Indo-Iranian-speaking + Dravidian-speaking ones. There is surely an ASI-related connection."

      I doubt that there is any Ancestral South Indian (ASI)-related component in Iranians and Kurds higher than noise. All connections to South Asia must be Ancestral North Indian (ANI)-related.

      Delete
  5. The number of tested individuals in both studies are too small to make bold assumptions. Let's take the sum:

    Grugni et al.: 57 Iranian Arabs
    3 individuals with R1b, 2 individuals with R1a

    Nasidze et al.: 46 Iranian Arabs
    2 individuals with R1b (or R2!), 5 individuals with R1a

    Sum: 103 Iranian Arabs
    Max. 5 individuals with R1b, 7 individuals with R1a
    R1b: max. 5%
    R1a: 7%


    Indeed, more sampling and study of Iranian Arabs is needed to settle the R1a/R1b issue in Iranian Arabs. Just like Persians, they may have some significant regional variation.

    The Gilaki people are from the Gilan province. The Gilan province is not in Western Iran as you stated but Northern Central Iran, not far away from the capital Tehran.

    By western Iran I mean the part of Iran west of the Dasht-e Kavir and Dasht-e Lut deserts. These two deserts are the most influential natural barriers within the territory of the Iranian mainland.

    Furthermore, you stated that Iranians [i]"have more R1a than R1b, which is the opposite of the West Asian norm"[/i]. Well, Lur and Gilaki are obviously Iranians and show less R1a than R1b, so either your statement at the beginning of our conversation is incorrect and/or all of these studies with N<100 for each population should be taken with a grain of salt. I prefer the later option.

    Note that I later made a modification to that statement by saying "the presence of more R1b than R1a in them and two Persian samples from western Iran is indicative of the R1b-R1a line (roughly speaking) passing through western Iran rather than Iraq". Kurds west of modern-day Iran break that rule by possessing more R1a than R1b, but Kurds expanded to most of today's Kurdish-majority territories outside modern-day Iran only during the Islamic era.

    I wrote that R1b/R1a difference is mostly along the different religious lines with Muslims of West Asia having more R1a than R1b, and Jews and (former) Christians having more R1b than R1a.

    All West Asian Muslim groups outside Arabia and Iran possess more R1b than R1a except the Kurds living there.

    I doubt that there is any Ancestral South Indian (ASI)-related component in Iranians and Kurds higher than noise. All connections to South Asia must be Ancestral North Indian (ANI)-related.

    Your evidence?

    ReplyDelete
  6. "Indeed, more sampling and study of Iranian Arabs is needed to settle the R1a/R1b issue in Iranian Arabs. Just like Persians, they may have some significant regional variation."

    Agreed.

    "By western Iran I mean the part of Iran west of the Dasht-e Kavir and Dasht-e Lut deserts. These two deserts are the most influential natural barriers within the territory of the Iranian mainland."

    I see.

    "Note that I later made a modification to that statement by saying "the presence of more R1b than R1a in them and two Persian samples from western Iran is indicative of the R1b-R1a line (roughly speaking) passing through western Iran rather than Iraq". Kurds west of modern-day Iran break that rule by possessing more R1a than R1b, but Kurds expanded to most of today's Kurdish-majority territories outside modern-day Iran only during the Islamic era."

    Depending on the study Kurds also have more R1b than R1a,

    Nebel et al.: 99 Iraqi Kurds (Muslims)
    16 individuals with hg1 (R1b+R1+R2, excluding R1a) = 16%
    11 individuals with hg3(Eu19) (R1a) = 11%

    Stenersen et al.: 100 Iraqi Kurds
    13 individuals with R1b = 13%
    11 individuals with R1a = 11%

    Gokcumen et al.: 31 Kurds from Central Anatolia
    2 individuals with R1b = 6%
    1 individual with R1a = 3%

    "All West Asian Muslim groups outside Arabia and Iran possess more R1b than R1a except the Kurds living there."

    See above.

    "Your evidence?"
    Whenever a component like "Gedrosia" (Dodecad), "Balochistan" (Harappa) or "Persian" (DNA tribes) appears in a ADMIXTURE run with K>11, the South Asian component disappears in Kurds. This component is basically West Asian, not South Asian.

    ReplyDelete
  7. This comment has been removed by the author.

    ReplyDelete
  8. Depending on the study Kurds also have more R1b than R1a,

    Nebel et al.: 99 Iraqi Kurds (Muslims)
    16 individuals with hg1 (R1b+R1+R2, excluding R1a) = 16%
    11 individuals with hg3(Eu19) (R1a) = 11%

    Stenersen et al.: 100 Iraqi Kurds
    13 individuals with R1b = 13%
    11 individuals with R1a = 11%

    Gokcumen et al.: 31 Kurds from Central Anatolia
    2 individuals with R1b = 6%
    1 individual with R1a = 3%


    I am aware of those results. But the general pattern among western Kurds according to the totality of the relevant genetic studies is more R1a presence than R1b. The Kurdish sample of Gokcumen et al. is from a single village and thus can in no way be used as representative.

    Whenever a component like "Gedrosia" (Dodecad), "Balochistan" (Harappa) or "Persian" (DNA tribes) appears in a ADMIXTURE run with K>11, the South Asian component disappears in Kurds. This component is basically West Asian, not South Asian.

    It is because, while being mainly West Asian, those components also capture the ASI-related South Asian genes. Compare the Fst distances of Iranians and Kurds to the southernmost and most-ASI-descended South Asian populations with those of non-Iranic and non-Gypsy West Asians; you will see that Iranians and Kurds have less Fst distance to the most-ASI-descended South Asian populations than non-Iranic and non-Gypsy West Asian populations have. The ASI connection is obvious.

    ReplyDelete
    Replies
    1. "It is because, while being mainly West Asian, those components also capture the ASI-related South Asian genes. Compare the Fst distances of Iranians and Kurds to the southernmost and most-ASI-descended South Asian populations with those of non-Iranic and non-Gypsy West Asians; you will see that Iranians and Kurds have less Fst distance to the most-ASI-descended South Asian populations than non-Iranic and non-Gypsy West Asian populations have. The ASI connection is obvious. "

      Since Reich et al. 2009 it is known that all Indians are a mixture ANI and ASI; the ASI no longer exist in non-admixed form, but in various degrees of admixtures with ANI; the closest living population to the ASI are the Andaman Islanders. I don't have numbers for the Andaman Islanders but I highly doubt that they any closer to Kurds than to other Muslim West Asians, e.g. Turks or Yemenese.

      The most ASI-descended South Asian population in Dodecad Globe13 is:
      Paniya_Ch from Chaubey et al.

      Adjusted distance for Paniya_Ch Chaubey et al. using Dodecad Globe13:
      #70: 82.1 Turkish_Aydin_Hodoglugil et al.
      #71: 82.2 Yemenese Rasmussen et al.
      #72: 82.8 Turks Behar et al.
      #73: 83.2 Lebanese Behar et al.
      #74: 83.4 Turkish_Kayseri Hodoglugil et al.
      #75: 83.5 Kumyks_Y Yunusbayev et al.
      #76: 83.6 Kurds_Y Yunusbayev et al.
      #78: 83.7 Kurd_D Dodecad
      #79: 83.8 Syrians Behar et al.

      Kurds are not closer to South Indian Paniya people (most ASI-descended South Asian populations) than Turks, Lebanese, Syrians or Yemenese.

      Any evidence for your claim that Kurds have ASI? Show me numbers.

      Delete
    2. You almost certainly haven't checked the Fst distance results. If you did, you would not make any of those claims about Kurds and would agree with me. If you provide your email address, I will send you an Excel file showing the Fst distances.

      BTW, Turks' genetic pull towards South Asians is mainly due to the small but by West Asian standards not so small Mongoloid admixture in Turks (as ASI is genetically closer to Mongoloids than to Caucasoids) rather than any ASI admixture.

      Delete
    3. This comment has been removed by the author.

      Delete
  9. This comment has been removed by the author.

    ReplyDelete
  10. BTW, for the results I presented above I actually used Globe13 Fst divergences of the 13 components.
    https://docs.google.com/spreadsheet/ccc?key=0ArAJcY18g2GadF9CLUJnTUdSbkVJaDR2UkRtUE9kaUE#gid=3

    I combined these Fst results with the actual results of each population of the Globe13 run.
    https://docs.google.com/spreadsheet/ccc?key=0ArAJcY18g2GadF9CLUJnTUdSbkVJaDR2UkRtUE9kaUE#gid=2

    http://dodecad.blogspot.com/2012/10/globe13-participant-results.html

    My method to calculate "adjusted distances" is now been used by the Harappa Ancestry Project as well. Zack from the Harappa Ancestry Project calls it "weighted distances".
    http://www.harappadna.org/2012/03/ref3-admixture-dendrograms/
    http://www.harappadna.org/?s=weighted

    ReplyDelete
  11. Thanks for the file. The most most-ASI-descended South Asian population in your data set are the Paniya, similar to the Dodecad list of populations.
    From your data the Fst distances of Paniya, India (south) PNYD3:
    ...
    #57 0.0778 Turkey tur37
    #58 0.0783 iran20
    #59 0.0819 Jordan560
    #60 0.0825 Syrians 485
    #61 0.0844 Adygei (Caucasus, Russia)
    #62 0.0847 Lebanon8
    #63 0.0847 Egypt12
    #64 0.0850 Lezgins Caucasus lez52
    #65 0.0853 North Russian (Arkhangelsk Oblast)
    #66 0.0854 Gypsies Romania7
    #67 0.0856 Yemen10
    #68 0.0860 Iraqi Kurd [F063888]
    #69 0.0869 Palestinian
    #70 0.0872 Stalskoe, Dagestan (Russia) [STAL35]
    #71 0.0876 Bedouin (Ngev, Israel)
    #72 0.0877 Armenians arm26

    This fits exactly to what I calculated and said. Kurds are not closer to South Indian Paniya people (most ASI-descended South Asian populations) than Turks, Lebanese, Syrians or Yemenese.

    ReplyDelete
    Replies
    1. Those Paniyas are the Behar et al. Paniyas, and as Zack demonstrated, most of the Behar et al. Paniyas are not Paniya but probably people from Southeast Asia thus Mongoloid rather than South Asian. So you should use a different population from the file I sent to you as a high ASI representative.

      Delete
    2. I see. For my previous Dodecad Globe13 calculation I did not use the problematic Behar et al. Paniyas but Chaubey et al. Paniyas. Unfortunately, your data set does not have Chaubey et al. Paniyas.

      The most-ASI-descended South Asian population in your data set is North Kannadi, India (south) KNTK388.
      From your data the Fst distances of North Kannadi, India (south) KNTK388:
      ...
      #21 0.0419 iran20
      #22 0.0424 Gypsies Romania7
      #23 0.0437 Kyrgyzstani [F063448]
      #24 0.0456 Turkey tur37
      #25 0.0469 chuvash17
      #26 0.0479 Jordan560
      #27 0.0486 Stalskoe, Dagestan (Russia) [STAL35]
      #28 0.0487 Lezgins Caucasus lez52
      #29 0.0489 Syrians 485
      #30 0.0489 Iraqi Kurd [F063888]
      #31 0.0499 Adygei (Caucasus, Russia)
      #32 0.0499 Mexican (California, USA) [Child]
      #33 0.0502 Armenians arm26

      Kurds are not closer to South Indian North Kannadi(most ASI-descended South Asian population in your data-set) than Turks, Caucasus people, Syrians or Jordanians.

      The second most-ASI-descended South Asian population in your data set is Mala, India (south; A.P.) [ML9]. They seem to have more ANI ancestry than North Kannadi.
      From your data the Fst distances of Mala, India (south; A.P.) [ML9]:
      ...
      #23 0.0357 iran20
      #24 0.0384 Turkey tur37
      #25 0.0393 chuvash17
      #26 0.0411 Jordan560
      #27 0.0412 Iraqi Kurd [F063888]
      #28 0.0412 Stalskoe, Dagestan (Russia) [STAL35]
      #29 0.0414 Lezgins Caucasus lez52
      #30 0.0416 Adygei (Caucasus, Russia)
      #31 0.0417 Syrians 485
      #32 0.0423 Mexican (California, USA) [Child]
      #33 0.0429 Mexican (California, USA) [Unrelated]
      #34 0.0436 Armenians arm26

      Kurds are not closer to South Indian Mala (second most ASI-descended South Asian population in your data-set) than Turks, Caucasus people or Jordanians.



      The third most-ASI-descended South Asian population in your data set is Madiga, India (south; A.P.) [M7]. They seem to have much more ANI ancestry than North Kannadi.
      From your data the Fst distances of Madiga, India (south; A.P.) [M7]:
      ...
      #23 0.0334 iran20
      #24 0.0358 Turkey tur37
      #25 0.0385 Iraqi Kurd [F063888]
      #26 0.0391 chuvash17
      #27 0.0392 Jordan560
      #28 0.0392 Stalskoe, Dagestan (Russia) [STAL35]
      #29 0.0395 Syrians 485
      #30 0.0398 Lezgins Caucasus lez52
      #31 0.0404 Adygei (Caucasus, Russia)
      #32 0.0411 Mexican (California, USA) [Child]
      #33 0.0412 Armenians arm26

      Kurds are not closer to South Indian Madiga (third most ASI-descended South Asian population in your data-set) than Turks.

      Delete
    3. No, the most-ASI-descended population in that file is Irula (there is no Pulliyar sample in that file):

      https://docs.google.com/spreadsheet/ccc?key=0AuW3R0Ys-P4HdFN6QlFNUWJYZVl4YmZUU3BIQ0NBeFE#gid=0

      Delete
    4. Okay, in your data set Irula are the most-ASI-descended population.

      From your data the Fst distances of Irula, India (south; T.N.) [I9]:
      #22 0.0569 iran20
      #23 0.0603 Turkey tur37
      #24 0.0622 Iraqi Kurd [F063888]
      #25 0.0626 Jordan560
      #26 0.0627 chuvash17
      #28 0.0631 Syrians 485
      #31 0.0645 Adygei (Caucasus, Russia)
      #32 0.0645 Stalskoe, Dagestan (Russia) [STAL35]
      #33 0.0646 Armenians arm26
      #34 0.0647 Lezgins Caucasus lez52
      #35 0.0671 Palestinian

      Agains, Kurds are in the range of Jordanians, Turks and Syrians.

      Zack wrote about the ASI component:
      "The South Indian component is about equally far from Caucasian, Baloch, NE Asian, NE European and SE Asian components. "
      http://www.harappadna.org/2012/06/harappaworld-ancestral-south-indian/#comments

      The Behar Iranians are not ideal because several individuals in this data set are clear outliers.
      http://www.harappadna.org/2011/03/iranians/

      This is why you will always see Behar Iranians being closer to Africans and South Asians when compared to other Iranian samples.

      Unfortunately, Xing et al data (e.g. the Iraqi Kurds in your data set) have very little overlap with other data, they are not good for comparisons, I explained this earlier:
      http://kurdishdna.blogspot.com/2012/07/some-clarification-for-xing-et-al-data.html

      In order to use them Zack just used 29,000 SNPs and was not happy about the noisy results.

      Unfortunately, your data set is based on a even smaller number of SNPs (13,600 SNPs), which makes signals even more noisy. FYI, if I recall it correctly then the Dodecad Globe13 data are based on 170,000 SNPs.

      Unfortunately, your data set does not have Chaubey et al. Paniyas nor the Metspalu et al. Pulliyas for further comparison.

      I presented adjusted distances of Paniyas earlier (using Dodecad Globe13), here are the adjusted distances of Pulliyas.
      ...
      #64 74.9 Iranians Behar et al.
      #65 75.7 Nogais_Y Yunusbayev et al.
      #66 75.9 Iranian_D Dodecad
      #67 77.4 Turkish_Aydin_Ho Hodoglugil et al.
      #70 77.7 Yemenese Rasmussen et al.
      #71 77.8 Turks Behar et al.
      #72 78.2 Kurds_Y Yunusbayev et al.
      #73 78.4 Lebanese Behar et al.
      #74 78.4 Turkish_Kayseri_Ho Hodoglugil et al.
      #75 78.4 Kurd_D Dodecad
      #76 78.5 Kumyks_Y Yunusbayev et al.
      #77 78.7 Kyrgyz_Bishkek_Ho Pagani et al.
      #78 78.9 Syrians Behar et al.
      #79 79.0 Turkish_D Dodecad
      #80 79.0 Turkish_Istanbul_Ho Hodoglugil et al.
      #81 79.2 Uzbekistan_Jews Behar et al.
      #82 80.0 Jordanians Behar et al.

      Kurds are not closer to South Indian Metspalu et al. Pulliyas (Most ASI-descended South Asian population) than Turks, Yemenese, Nogais. The distance to Pulliyas is somewhat like Lebanese.


      I think I now understand your thought process:
      you are saying that Iranians show some lower Fst divergences towards South Asia, and Iranians are closest to Kurds, so Kurds have some of the Ancestral South Indian component.

      The data do not fit though.

      Delete
    5. Okay, in your data set Irula are the most-ASI-descended population.

      It is not my dataset, but a friend's.

      Agains, Kurds are in the range of Jordanians, Turks and Syrians.

      I already made an explanation about Turks' results:

      "Turks' genetic pull towards South Asians is mainly due to the small but by West Asian standards not so small Mongoloid admixture in Turks (as ASI is genetically closer to Mongoloids than to Caucasoids) rather than any ASI admixture."

      Zack wrote about the ASI component:
      "The South Indian component is about equally far from Caucasian, Baloch, NE Asian, NE European and SE Asian components. "
      http://www.harappadna.org/2012/06/harappaworld-ancestral-south-indian/#comments


      The "South Indian" (or "South Asian") component of Zack's analyses is not full or close to full ASI, it is in Zack's words:

      "South Indian component is a mixture of ASI (major) and ANI (minor)".

      ASI is genetically closer to Mongoloids than to Caucasoids.

      The Behar Iranians are not ideal because several individuals in this data set are clear outliers.
      http://www.harappadna.org/2011/03/iranians/

      This is why you will always see Behar Iranians being closer to Africans and South Asians when compared to other Iranian samples.


      There is not much difference between the average genetic results of the Behar and the other Iranian samples.

      I think I now understand your thought process:
      you are saying that Iranians show some lower Fst divergences towards South Asia, and Iranians are closest to Kurds, so Kurds have some of the Ancestral South Indian component.


      Kurds do have ASI more than noise levels, but less than Iranians.

      Delete
    6. Your almost obssesive mentioning of this ASI component makes me wonder.

      The ASI in Kurds is noise. The genes connecting Kurds to South Asia are almost entirely ANI. Dienekes had several runs, articles about that. in the higher Ks with more components. The ANI component of South_Central Asia is absorbed by Gedrosia. But than Gedrosia itself is the closest what comes to original West Asian. Last year Dienekes published a break up of Components in his block in which it was obvious that the South Asian absorbed by Gedrosia is ANI not ASI. Alone from the Fst distances this get obvious. The ASI component in Kurds is actually even lower than among Turks or populations of the Levant.

      Delete
  12. what are the genetic diferences from people of dersim and other kurds?
    cultural they are different.

    ReplyDelete
    Replies
    1. We have automsomal genetic information of three individuals from Dersim, one Zaza speaker (KD024) and two Kurmanji speaker (KD002, KD004), by religion they are all Alevis.
      http://kurdishdna.blogspot.com/2012/06/kurdish-autosomal-dna-based-on-dodecad.html

      Why do think that people of Dersim are culturally different from other kurds?
      Religion: You can find Alevis in a lot of areas outside of Dersim.
      Language: You can find Zaza and Kurmanji speakers in a lot of areas outside of Dersim.

      Delete
  13. in Turkey all have this feeling even kurds, the old PKK leader abdullah Öcalan had a famous definition "Dersim caracter"

    yes there are other alevis and zazas but they are different too.

    ReplyDelete
  14. This comment has been removed by the author.

    ReplyDelete
  15. @Kurti,

    You are reiterating the arguments that I already refuted above. Kurds possess ASI ancestry more than Turks and more than noise levels but less than Iranians. This is obvious to everyone who knows enough about the genetics of these peoples and treats them with an unbiased and objective mind. Is it so hard to acknowledge this simple fact?

    ReplyDelete