Gypsy DNA

[Stevo]

What do we know about the Gypsy or Roma people and their DNA?

A little can be found on the ISOGG site here.

It mentions y-haplogroup H:

Y-DNA haplogroup H: The founding mutation for haplogroup H, M69, occurred in a haplogroup F man, probably in the Indian subcontinent The founder of haplogroup H probably lived about 30,000-40,000 years ago. This haplogroup has not yet been studied in a comprehensive manner. Today, nearly all members of haplogroup H live in the Indian subcontinent area. The Roma (also known as Gypsy) people, who apparently originated in India, are the main source of haplogroup H in western Europe.

In googling "Gypsy DNA" I also saw mention of y-haplogroup R2.

Anyone know any more about this subject?

[Stevo]

Here is an interesting site with photos of modern gypsies in England.

At the bottom right you can click on a link to photos of gypsies in other parts of Europe.

[Victor]

Just a couple of additional references not listed in ISOGG's page

Origins and Divergence of the Roma (Gypsies)

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1235543

Patterns of inter- and intra-group genetic diversity in
the Vlax Roma as revealed by Y chromosome and
mitochondrial DNA lineages

http://hpgl.stanford.edu/publications/EJHG_2001_v9_p97.pdf

There are a couple of other studies related to the Romani but I just couldn't find the full text versions.

The haplogroup that I've also seen associated with the Roma people, besides H, is R1a. Check some of the tables at the following link.

http://wanclik.free.fr/adn.htm

Victor

[cacio]

Two articles I know of are the following:

Gresham et al: Origin and divergence of the Roma
Kalaydjieva et al: Patterns of inter- and intra-group genetic diversity in the Vlax Roma

Gresham (with a sample mostly from Bulgaria) finds the following for the Y chromosome (but this has to be verified, they use a non-standard notation that I tried to map):
H1 44%
I 22.6%
J2f 12.6%
R1 6.7%
FxI,J,H1,R 3.6%
E3b 3.6%

Kalaydjieva (even more difficult to read) finds 73% H in her Vlax sample,
with the second largest being (I believe) FxJHK.

cacio

[cacio]

Sorry Victor, I didn't see your post before I submitted.

Looking at the data, one would think that the Indian component is essentially only H. The other major Indian groups (R1a, R2, L) seem small or absent,
and this even more after accounting for the fact that Eastern Europe has a lot of R1as. J's are present in India but could have been picked up in the Middle East.

What puzzles me is Haplogroup I. Haplogroup I is very non Indian. Could it be sign of an admixture in the Balkans? I is found also (5/20) in their smaller Spanish Roma sample.

I didn't write down the mtdna data. Lots of M's (Indian), but also many H, U etc.

cacio

[Jim Honeychuck]

Gypsy DNA

Last Friday DNA Tribes posted a sample autosomal genetic profile:

http://dnatribes.com/sampleresults.html

Jim

[Stevo]

Sorry Victor, I didn't see your post before I submitted.

Looking at the data, one would think that the Indian component is essentially only H. The other major Indian groups (R1a, R2, L) seem small or absent,
and this even more after accounting for the fact that Eastern Europe has a lot of R1as. J's are present in India but could have been picked up in the Middle East.

What puzzles me is Haplogroup I. Haplogroup I is very non Indian. Could it be sign of an admixture in the Balkans? I is found also (5/20) in their smaller Spanish Roma sample.

I didn't write down the mtdna data. Lots of M's (Indian), but also many H, U etc.

cacio

It certainly seems that H is the defining y-haplogroup among the Romani (gypsy) people.

I would like to know more about the Y-hg I found among them, as well. Is it the undifferentiated I of the Middle East and the Caucasus, the East Euro I1b of the Balkans, or some other variety of I? I didn't get that info from skimming those articles.

It does seem that there is much less R among the Romani than in the Indian population in general.

[Marttinen]

I remember a recent article I read that placed Macedonian Roma as 44% E3b1. This was the greatest percentage of E3b1's in the Balkans (maybe even in Europe) outside of the Kosovar Albanians.

Now, about mtDNA. Would it be a good chance that someone who is from an established family in Europe (in this case, France) and is Haplogroup M* would have a genetic connection to the Roma?

[Kaiser]

The Indian YDNA 'H' and mtDNA 'M' dominate amongst the Roma, at nearly 50% and 25% respectively. Additionally, there is a smattering of I, J, & R sub-clades. Linguistic affinity with the Punjabis cannot go unnoticed as many words are comprehensible in meaning and pronunciation.

A sub-group of the Roma is known as Sinte (or Sinti) who had moved from their original homeland in South Asia to erstwhile East Germany from where they were re-located to Uzbekistan by the Communist regime. The Sinti males have R2 as the dominant haplogroup at a high frequency of 50%. Like H, R2 is an indigenous South Asian YDNA haplogroup. (Outside South Asia, R2 has spikes amongst Chechens and Kurdish populations.)

It seems that each of the two dominant Indian male haplogroups ie, H & R2 are well represented amongst the 'wandering nomads' who can be seen as a minor third category, in addition to castes and tribes. In South Asia, the gypsies are known as Banjara, Changar, Odh, etc and their social behaviour is pretty much the same as that of European Roma.

[cacio]

Stevo:

Gresham only tests for M170, not subgroups. He does provide a set of 8 STR markers, so may be it's possible to infer from those, but I'm not familiar with this type of analysis. Kalaydjeva tests even less; but she also provides some STRs.

It seems to me reasonable to assume that Y hap H and mtdna M in Europe have a high probability of being gypsy. One could have hypothesized the same of my own haplogroup L, but so far L has not been found among the gypsies. (Moreover, while L is present at appreciable percentages from India to Anatolia, H is not.) From these previous two posts it seems though that gypsy groups may be more heterogeneous than what the Gresham and Kalaydjeva paper seem to imply.

cacio

[Kaiser]

Martinnen --- Presence of mtDNA 'M' in an "established European (French) family" would surely qualify for an Indian connection, but given the close-knit endogamous Roma communites, is quite improbable. Elopement of a pretty gypsy girl with a French nobleman could, however, be a rare happenstance. But have you heard of such liasions that you have referred to?

[Victor]

I remember a recent article I read that placed Macedonian Roma as 44% E3b1. This was the greatest percentage of E3b1's in the Balkans (maybe even in Europe) outside of the Kosovar Albanians.

Now, about mtDNA. Would it be a good chance that someone who is from an established family in Europe (in this case, France) and is Haplogroup M* would have a genetic connection to the Roma?

That would be an interesting finding. Please read the abstract below which seems to show a different picture.

Forensic Sci Int. 2005 Nov 25;154(2-3):257-261.

Population genetics of 8 Y chromosome STR loci in Macedonians and Macedonian Romani (Gypsy).

Pericic M, Klaric IM, Lauc LB, Janicijevic B, Dordevic D, Efremovska L, Rudan P.

Institute for Anthropological Research, Amruseva 8, 10000 Zagreb, Croatia.

Eight Y chromosome short tandem repeat (STR) polymorphisms (DYS19, DYS385, DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393) were analyzed in Macedonians (n=84) and Macedonian Romani ethnic group (n=68). Observed allelic frequency distribution and locus diversity values in Macedonians correspond closer to neighboring southeastern European populations than (mostly) western European populations, whereas observed allelic frequency distribution and locus diversity values in Macedonian Romani, as expected based on their Asian (Indian) origin, differ from both neighboring southeastern and (mostly) western European populations. Sixty-six (78.57%) haplotypes appeared in single copies in Macedonians and 15 (22.06%) in Macedonian Romani. The most frequent Macedonian haplotypes (DYS19-DYS385-DYS389I-DYS389II-DYS390-DYS391-DYS392-DYS393) 16-14/15-13-31-24-11-11-13 and 13-16/18-13-30-24-10-11-13 were found in 7 and 6 copies, respectively. The most frequent Macedonian Romani haplotype 15-15/17-14-29-22-10-11-12 was found in 18 males. Total haplotype diversity was 0.9885+/-0.0058 (Macedonians) and 0.9008+/-0.0242 (Macedonian Romani).

PMID: 16182975 [PubMed - as supplied by publisher]

If you noticed Sixty-six (78.57%) haplotypes appeared in single copies in Macedonians and 15 (22.06%) in Macedonian Romani. Running the haplotypes in the abstract thru Whit's predictor, the 78 percent is comprised by E3b and I1b while the 22% is comprised by H. The 78% are simply Macedonians and the remaining 22% are the Macedonian Romani.

If you find the article you refer to please let us know.


p.s. Sorry that the full text article of the abstract above is not freely available online.

[Marttinen]

The source is here: http://vetinari.sitesled.com/slavic.pdf

It seems though, that between my memory and reading the table wrong (it was a fair distance across the screen that my eyes had to follow) my percentages were off.

Macedonian Romani: 29.8% Albanian Kosovar: 45.5%

It did (Table 1) show, however, that these were the two groups with the highest concentration of E3b1 in Europe.

[Stevo]

The source is here: http://vetinari.sitesled.com/slavic.pdf

It seems though, that between my memory and reading the table wrong (it was a fair distance across the screen that my eyes had to follow) my percentages were off.

Macedonian Romani: 29.8% Albanian Kosovar: 45.5%

It did (Table 1) show, however, that these were the two groups with the highest concentration of E3b1 in Europe.

What caught my eye in that Pericic, et al, study, as an R1b1c guy, was the comment that 43% of Ossetian males are R1b.

I wasn't aware of that.

I know the Ossetians are not gypsies, and gypsies are the topic of this thread.

I was just surprised to read that.

The Ossetians are supposed to be the descendants of the Alans, Sarmatians, and Scythians.

Interesting.

[Marttinen]

About the M*

The person I am interested in was adopted, like me. She was told her mother is French and was supplied with a rather rare French last name. Googling the name gave results (business names, tombstone inscriptions, marriage certificates) around the island of Jersey (which is close to France, anyway) and a couple of scattered areas in Canada (the country this female lives).

Anyway, I'm trying to figure out how my wife with a French/Italian background ended up with M* mtDNA. She had a couple low-grade (never tested for high grade) matches with people from Spain, the Azores and USA.

I tried to look for studies of M* percentages in Europe but have come up blank. There are some M1's (and 3 of her low-grade matches were M1's) but M* is always mentioned in connection with the Indian sub-continent.