Possible Romany Ancestor?

[Alloro810]

Hello,

I am wondering if I have a possible Romany ancestor and I would like the forum's take on this.

I just came across a thread the other day in another discussion forum and someone had mentioned that as long as South Asian DNA appears in your results, even if it's one percent can indicate gypsy heritage (provided that you have no other South Asian connections).

I was curious to see if I did have this in my background, and to get your take on it. I had my DNA tested through FTDNA, and having an analysis by Dr. J. MacDonald, there appeared a small amount of South Asian DNA.

I have no known connection to South Asia at all. I do have significant Middle Eastern DNA (my mother's side of the family is Southern Italian). At first, I theorized that the only way the South Asian DNA came in was through someone of Arab and South Asian descent. However, I am not so sure of that.

Upon further analysis of my results by the Dodecad Ancestry Project calculators, I see 0.01% South Asian DNA using the dv3 calculator. With the Dodecad Oracle calculator, out 30 reference populations, the top three results were Tuscans and Romanians.

Another calculator, the eurasia7, lists the South Asian percentage as 0.48%. A third calculator, world9, lists it again as 0.01%. So, in total, two Dodecad calculators list it as 0.01% and one lists it as 0.48%.

Does a percentage total such as 0.01% or 0.48% of South Asian DNA, no known connection to South Asia (but a clear DNA connection to Western Asia) indicate a Romany ancestor somewhere? My known ethnic heritage is Italian, Irish and Polish. I am not sure what to make of all this.

[thetick]

You asked "Does a percentage total such as 0.01% or 0.48% of South Asian DNA, no known connection to South Asia (but a clear DNA connection to Western Asia) indicate a Romany ancestor somewhere? "

Not at all. Any percentage below 2% is likely noise. 1% could be real if you have paper trail but it's more likely noise. Not sur who told you 1% is real, since more likely it's not. Why would think you have Romany or South Asian ancestry?

[Alloro810]

No one told me that the percentages I received were real - I read a posting from another forum that mentioned a similar statement, so I was simply asking for clarification. South Asian DNA does come up in my results and having further analysis shows a possible Romanian link. From what I have read, the Roma are very much present in Eastern Europe, particularly Romania. I have Eastern European ancestry already (Polish) and it is possible there may be Romanian in there. I don't know too much about that branch of the family and I was curious to see, based upon the DNA results and what I know of my background, if there was a Roma component to the mix. Just curious. It's ok if there's not. Any information to help genealogical research is always good. Thanks for your response.

[Donald Locke]

I am the Admin of the Romany DNA project. I agree with thetick, that your percentage is so low it is likely noise and not representative of Romany blood ties.

I will use my own results as a comparrison. Running my FF through the
Dodecad and Harappa calculators shows very different results.
Harappa shows a much higher South Asian and South West Asian percentage then the Dodecad calculator does. I am more inclinded to go with the
Harappa calculation mainly because that study is focused on the South Asian population.

FTDNA FF population finder shows me as 100% European, but all other calculators have something else to say on the matter.
I wish FTDNA would impliment a better population finder more inline with the various calculators available out there. To tell me I am 100% European is not the whole truth. From what I have seen, the Harappa calculator for those with proven South Asian roots, has a more consistent calculator I think.


My Dodecad results

8.18% East_European
57.22% West_European
22.36% Mediterranean
8.42% West_Asian
1.99% South_Asian
0.56% Northeast_Asian
0.23% Southwest_Asian
1.04% Palaeo_African


My Harappa results

7% South Asia
2% Onge
21% South West Asia
66% European
1% Siberian
1% Paupan
2% American

As you see the Harappa results tell a much different story then the Dodecad calculator. I think this is because the Harappa calculator is more inclinded to compare South Asians then the Dodecad calculator.


Here are the Harappa calcuations on other Romany results for folks to compare to.

6% South Asia
21% South West Asia
70% European
1% American
1% East Africa
-------

20% South Asia
6% Onge
1% East Asia
31% South West Asia
40% European
1% Papuan
1% American
------

13% South Asia
3% Onge
20% South West Asia
64% European
-----

13% South Asia
4% Onge
18% South West Asia
65% European

-------

9% South Asian
3% Onge
17% South West Asia
66% European
1% Siberian
2% Papuan
1% American
1% San/Pygmy
1% East African

------

8% South Asia
4% Onge
19% South West Asia
54% European
1% Siberian
1% Papuan
7% American
1% San/Pygmy
6% East African

[Donald Locke]

While there are Romany of Romania, obviously not all Romanians are Romany.
Romany represent a small portion of the Romanian population.
Poor Romanians are not always Romany in origin, but are often mistaken as Romany when they are not. There is a very real distinction between a
Romanian Romany, and a Romanian, Romany are clearly South Asians in origin.

So those of you with Romanian roots, understand that does not imply Romany roots, Romany just happen to live in Romania, but that Romany and Romanians are not one and the same people. Romany live in just about every European nation today and should not be confused with Europeans and European Travelers.

Through the Romany DNA project and Y Haplo Group H project, I have many multiple Romany of Europe Y DNA tested, and can safely say that around 70% of the European Romany male population carries Y Haplo Group H1a with a very specific null value marker mutation, marker 425 = 0.

Romany of England, Scotland, Bulgaria / Moldova, Hungry, Russia have all been identified as Y Haplo Group H1a men with the 425 = 0 marker mutation.
This null value marker mutation so far has not been found in any H1a Asian male population, this null mutation appears to be distinct and unique to the European Romany H1a population.

If any of you can prove you have an Y Haplo Group H1a with the 425 = 0 marker mutation in your family tree, would be your Romany connection in your family tree.

I have clearly proven some of the Colonial American male lineages do in fact carry Y Haplo Group H1a with the 425 = 0 marker mutation from the 1600's - 1700's era, Maryland, Virginia and South Carolina.
And that fits the history! England and Scotland both were forcefully transporting Romanichal convicts to the American Colonies as early as the 1670's if not earlier. Yet there is virtually no mention of South Asian Indians / Romany in the history text books.

DNA does not lie, many multiple Y Haplo H1a with the 425 = 0 marker mutation male lineages of Maryland, Virginia, South Carolina has been identifed by me, and all our trees trace back to those 3 specific colonies which agrees with the convict records that very clearly state Romanichal "Gypsy" convicts were transported to Maryland, Virginia, South Carolina,
New York. How many actually made it to New York is unknown, many were likely dropped off in Southern Colonies.

The convict records state they were being shipped to New York too, but how many actually made it to New York is impossible to know with out some intensive paper trail research. But they certainly made it to Maryland, Virginia and South Carolina by the 1600's - 1700's era.

Lock / Locke, Ingram, Bailey / Baillie, Campbell, Hite, Carter, Ruffin, Short are all Colonial American male lineages found carrying Y Haplo Group H1a, which is clear evidence there were in fact Romany in the American Colonies far earlier then most historians openly acknowledge.

Further DNA testing done on English Romanichal male lineages, proves the following: Smith, Lock(e), Stanley, Boswell, Burton, Lee male lineages also carry Y Haplo Group H1a with the 425 = 0 marker mutation.

The English Romanichal's share the same Y chromosome with their Eastern European Romany cousins. The Romany are dominately descended from a single paternal forefather with in the last 2000 years, since the time the Romany left Asia for Europe.
Greenspan agrees that the Y Haplo H1a Romany male lineages all shared a single paternal forefather with one another with in the last 2000 years, while in Europe.

Finding a Y Haplo Group H1a male lineage in your tree, would very likely be your direct Romany blood tie.
mt Haplo Group M5a1 while not specific to the Romany population because it is South Asian in origin, if you are an M5a1 European female could also be very highly suggestive of your Romany roots.

More research is needing to be done on the M5a1 female lineages of Europe.
I know far more about the male Haplo H1a then the female M5a1 at this time.

Don

[Donald Locke]

I didn't personally run my Dodecad results, a cousin ran it for me.
But as I understand it, you download your FF autosomal results and upload them in to the Dodecad calculator. But exactly how it works I don't know because I didn't do it myself.

[briancowings]

Gedmatch.com has a utility that uses the same dataset as DIYDodecad V2.

'Admixture Proportions'

"This utility uses the same dataset as DIYDodecad V2. Different software is being used to enable it to work on this web site.
Even though the reference population names are the same, the results produced by this utility will be less accurate that those produced by DIYDodecad V2 primarily because it is limited to 1000 iterations to limit server CPU use.
Credit goes to Dienekes Pontikos for his excellent work."

I am asking help to send my results to Dodecad Calculator, but nobody help me. Can you help me?

[Alloro810]

Harappa calculator

I am the Admin of the Romany DNA project. I agree with thetick, that your percentage is so low it is likely noise and not representative of Romany blood ties.

I will use my own results as a comparrison. Running my FF through the
Dodecad and Harappa calculators shows very different results.
Harappa shows a much higher South Asian and South West Asian percentage then the Dodecad calculator does. I am more inclinded to go with the
Harappa calculation mainly because that study is focused on the South Asian population.

FTDNA FF population finder shows me as 100% European, but all other calculators have something else to say on the matter.
I wish FTDNA would impliment a better population finder more inline with the various calculators available out there. To tell me I am 100% European is not the whole truth. From what I have seen, the Harappa calculator for those with proven South Asian roots, has a more consistent calculator I think.


My Dodecad results

8.18% East_European
57.22% West_European
22.36% Mediterranean
8.42% West_Asian
1.99% South_Asian
0.56% Northeast_Asian
0.23% Southwest_Asian
1.04% Palaeo_African


My Harappa results

7% South Asia
2% Onge
21% South West Asia
66% European
1% Siberian
1% Paupan
2% American

As you see the Harappa results tell a much different story then the Dodecad calculator. I think this is because the Harappa calculator is more inclinded to compare South Asians then the Dodecad calculator.


Here are the Harappa calcuations on other Romany results for folks to compare to.

6% South Asia
21% South West Asia
70% European
1% American
1% East Africa
-------

20% South Asia
6% Onge
1% East Asia
31% South West Asia
40% European
1% Papuan
1% American
------

13% South Asia
3% Onge
20% South West Asia
64% European
-----

13% South Asia
4% Onge
18% South West Asia
65% European

-------

9% South Asian
3% Onge
17% South West Asia
66% European
1% Siberian
2% Papuan
1% American
1% San/Pygmy
1% East African

------

8% South Asia
4% Onge
19% South West Asia
54% European
1% Siberian
1% Papuan
7% American
1% San/Pygmy
6% East African

I have never heard of the Harappa calculator before so this is very interesting. I looked through the site for the project but I wasn't able to see if it's possible to download the calculator like I did with Dodecad. Can you download this calculator yourself or do you have to be accepted into the project in order to have your admixture results analyzed?

[thetick]

Currently you need to be a member of the project to get your results. The author has stated in his latest blog post "I am working on a new Admixture calculator." see http://www.harappadna.org

I suspect in a couple of weeks we will all be able to run a Harappa Ancestry Project calculator.

In the mean time there are plenty of calculators you can run from Dodecad, Eurogenes and MDL Projects.

[Alloro810]

Admixture calculators

I have used Dodecad, but not the other two. What do you think of them?

[irenych]

Any percentage below 2% is likely noise. 1% could be real if you have paper trail but it's more likely noise.

Just a technical question. I have heard this expression used frequently. But what exactly is "noise"? What is that supposed to indicate?

[Alloro810]

Just a technical question. I have heard this expression used frequently. But what exactly is "noise"? What is that supposed to indicate?

Testing error or "noise" is I think margin of error. If there is a 5% margin of error, that means the result is either 5% less than what it is, or 5% more than what it is.

[irenych]

Testing error or "noise" is I think margin of error. If there is a 5% margin of error, that means the result is either 5% less than what it is, or 5% more than what it is.

So if you get 1% "South Asian" noise, that means you have either 2% South Asian or no South Asian?

[Alloro810]

Testing noise

No, if you get a result of 1% South Asian, that means the result is 1%. That 1% may be noise, it may not be. What the other poster was saying was that anything below 2% is most likely noise. One calculator told me that I have 0.01% South Asian ancestry. It is below 2% and such a low amount at that, meaning that it is most likely noise. I am going to test with another calculator to determine if this result is correct. It could also be true ancestry, but the best way to back that up is with extensive genealogical research and a paper trail, if you can.

[thetick]

In the general sense any noise is error involved in any scientific calculation.

Now specifically in these tests the biggest component of noise is probably that every sample population is slightly admixed. So the 1% you get with South Asian may indeed be 1% with South Asians, but it's likely that 1% may be actually be European in the South Asian samples.

I think these test provide some very interesting data. The problem is people get all hung up on small percentages which in most cases are meaningless.

Another factor that creates the noise: Note for example the 11% Mediterranean component in Northern Scotland? Just added proof the Mediterranean named component does not correspond with what every one thinks as modern day Mediterranean. The names that you think are familiar like Northern European etc are ANCIENT components that were named by the authors of the tests based on where geographically they are most common NOWDAYS.

See Dodecad component maps at http://bergep.clanteam.com/DOD_Maps/dod_map.html