PDA

View Full Version : Turkmen haplogroup Q = genetic drift?



Afshar
11-24-2016, 07:57 AM
So there is a new russian paper released where the Turkmen from Karakalpakstan have 74% Q-M25. http://www.anthrogenica.com/showthread.php?709-New-DNA-Papers&p=198965&viewfull=1#post198965
Cristofaro et al found 31.1% Q-M25 in Afghanistan Turkmen, and Grugni et al found 42.6% in Iranian Turkmen.
(I should also note that Balaresque et al (2015) has found no Q in Turkmen also from Karakalpakstan)
So now we have three different papers, with a majority of a single hg (Q-m25), spread over different regions, can this be the result of genetic drift?

Arame
11-24-2016, 08:03 AM
I wouldn't call it drift. It is most probably a founder effect created by a prominent clan.

Afshar
11-24-2016, 08:09 AM
Yes that came to my mind, except that its impossible all belong to the same tribe. Geographically its not a match.

DMXX
11-24-2016, 09:17 AM
It certainly is genetic drift, and it's also a founder effect. :P

Founder effects are a subset of genetic drift, as are bottlenecks. The main difference between founder effects and bottlenecks are circumstantial (former is a natural/voluntary outgrowth of a daughter population from a parent, latter is a reduction in allelic diversity due to environmental factors, such as natural disasters, war, infection etc.). Further reading with diagrams here (http://evolution.berkeley.edu/evolibrary/article/bottlenecks_01).

From memory, Y-DNA Q is also the most common East Eurasian Y-haplogroup among Azeris (both Republic and Iranian) as well. We're very likely looking at one of the Oghuz signals.

Would be immensely useful to obtain some of these samples for further testing.

Afshar
11-24-2016, 09:35 AM
I wouldn't call it drift. It is most probably a founder effect created by a prominent clan.

You are probably right. The paper estimates the TRMCA as around 1090 which coincides with the appearance of Oghuz tribes in the region. The ethnogenesis of different tribes we know started probably a while later, and so the Q could be from a prominent clan who were there in the beginning.


It certainly is genetic drift, and it's also a founder effect. :P

Founder effects are a subset of genetic drift, as are bottlenecks. The main difference between founder effects and bottlenecks are circumstantial (former is a natural/voluntary outgrowth of a daughter population from a parent, latter is a reduction in allelic diversity due to environmental factors, such as natural disasters, war, infection etc.). Further reading with diagrams here (http://evolution.berkeley.edu/evolibrary/article/bottlenecks_01).

From memory, Y-DNA Q is also the most common East Eurasian Y-haplogroup among Azeris (both Republic and Iranian) as well. We're very likely looking at one of the Oghuz signals.

Would be immensely useful to obtain some of these samples for further testing.

Need to read me in to these things again :).

Kristiina
11-24-2016, 10:54 AM
Q-M25 has been detected in BA and IA Altai in Karasuk culture (together with R1a1-Z2124) and during the Iron Age (together with J2a2 and R1a1-Z93 (R1a1-Z2123?) (these haplotypes are based on Genetiker's work). Today, R1a1-Z2123 is typical for such Turkic speakers as Bashkirs, Chuvash, Tatars, Karachay-Balkar, Tajiks and Azers. Do you know if there is R1a1-Z2123 or R1a1-Z2124 also in Turkmens?

In any case, the Turkic ethno genesis could have a connection with BA and IA Altai.

Afshar
11-24-2016, 11:01 AM
Q-M25 has been detected in BA and IA Altai in Karasuk culture (together with R1a1-Z2124) and during the Iron Age (together with J2a2 and R1a1-Z93 (R1a1-Z2123?) (these haplotypes are based on Genetiker's work). Today, R1a1-Z2123 is typical for such Turkic speakers as Bashkirs, Chuvash, Tatars, Karachay-Balkar, Tajiks and Azers. Do you know if there is R1a1-Z2123 or R1a1-Z2124 also in Turkmens?

In any case, the Turkic ethno genesis could have a connection with BA and IA Altai.
Yes the Karasuk connection is definitely there.
The Balaresque paper (http://www.nature.com/ejhg/journal/v23/n10/full/ejhg2014285a.html) has an abundance of R1a, but did not test for specific snps.

Arame
11-25-2016, 04:11 AM
Would be immensely useful to obtain some of these samples for further testing.

Raw data probably will be available after the publication in peer reviewed journal.
One of authors suggested.