MtDNA help needed please!!!

[burto]

Hi,
Just got Mum's DeCodeMe results back and they say her MtDNA is R* but Oxford Ancestors said H!!!
Are they from the same branch, because they don't appear to be on the maternal lineage diagram I have from DeCodeMe!!!

[Maria_W]

Happy she got results...

Burto,

Happy to hear that your mom got her results back... I hope that the ancestry portion gave you the results you were hoping for... I really don't know that much about mtDNA so can't really help you there... I am a H1a1.with 16 mutations. My oldest known female line goes back to Pennsylvania in 1715 with Eva Brenneman..(German). My maiden name Shook is also German..So I am a Heinz 57(variety) of nationalities. Hope someone can help you!

Maria

[cacio]

Do you have your HVR1 and/or HVR2 from Oxford ancestors results? R is the "ancestor" of H. Usually, the star means you belong to R but not to any of its known subgroups. But this would be very strange. There is really no R* in Europe, at most, there is HV (which is the mother of H, so to speak).

It's kind of hard to read decodeme mtdna data, but you can ask them as well to interpret how they gave this result. It could simply be a clerical error.

cacio

[rainbow]

Do you have your HVR1 and/or HVR2 from Oxford ancestors results? R is the "ancestor" of H. Usually, the star means you belong to R but not to any of its known subgroups. But this would be very strange. There is really no R* in Europe, at most, there is HV (which is the mother of H, so to speak).

It's kind of hard to read decodeme mtdna data, but you can ask them as well to interpret how they gave this result. It could simply be a clerical error.

cacio

Wow! That is one heck of a clerical if it is one! If I were you I email both comanies and tell them about the two completely different results! One of two companies gave you the wrong info. I hope you get a refund from the one that is wrong. I hope they recheck it. If they both say they are right then try another company. I would recommend the Genographic Project. Theirs is about one hundred dollars plus tax. Or FTDNA. I think I once read somewhe

[burto]

Hi,
I have emailed DeCodeMe and I'm just waiting for a reply. On the results it say that it is extremely rare for anyone of Western Eurasian or European ancestry to get R* and it is most common in South Asia.
If you look on their demo of the maternal lines, it shows that R* and H are both descended from R however they are 2 seperate branches.
We had a basic test done by Oxford Ancestors as they were doing a special offer so don't know the HVR1 etc it just came back as Helena, but according to the diagram on DeCodeMe Helena and R* are 2 seperate lines!
Although I don't want to jump to conclusions and read anything into this until I've heard back from DeCode, AbDNA and DNATribes both have acknowledged that my Mum has South Asian ancestry, however unless their was an adoption, I can only think it would have been an ancestor in the late 1700's to early 1800's that can be responsible as I can trace my direct maternal line back to 1830. Will have to wait and see! They will probably offer a retest and we will have to go through the whole waiting process again!!!

[burto]

Just got my reply back and they tell me that because of the fragment of MtDNA control region OA uses does not always provide adequate info to distinguish between closely related mitogroups like H and R* whereas they use mt SNPs specifically used to select between different mito groups so they are confident R* is correct. Have emailed them just to double check however I am stunned and it appears that it's my Mum's maternal family that have turned out to be my unknown family!!! How ironic that I'm searching for my unknown US ancestry and find something exotic in my British ancestry!! Ok off to look at mitosearch!!!

[burto]

Erm...doe anyone know how to put your mutations from DeCodeMe into mitosearch? The numbers don't seem to match? Is it HVR1 or HVR2?!

[cacio]

I think Ann Turner created a spreadsheet with the mapping from decodeme to the usual CRS locations, but I don't remember the link right now.

However, you cannot really create a user in mitosearch, because decodeme only tests a few locations. For instance, my version only tests 10 location in HVR1 (ie after 16000), while in reality there are 560! So you could be missing several mutations.

Btw, what's your status at decodeme 14767? Perhaps this could be indicative of H (ie C indicates H? but that's just a guess). Anyway, from what you've said, it is possible that decodeme doesn't test for the markers that are key for determining certain haplogroups, so R* may simply mean that you are somewhere in there, but the markers analyzed don't allow them to be more specific. I still keep thinking that R* in Europe is highly unlikely. Even in India, it may be present, but it's not that common either.

cacio

[burto]

Confused

So what you're saying is I'm not R*? Why would DeCodeMe analyze mtdna if they can't do it properly? Why do people market tests if they don't do what they say they do? They told me that they use more mt SNPs to determine different mitogroups?
I'm totally confused now. Have just wasted even more money on the wonderful "science" of DNA?

[burto]

Link to DeCodeMe's R* through their demo below looks completely seperate to H to me:
http://demo.decodeme.com/ancestry/female-line-info/R*
I can't even figure out what I thought is supposed to be the most reliable (MtDNA)
Taken from DeCodeMe:

Mitogroups in the deCODEme Complete Scan

The deCODEme Complete Scan uses 163 SNPs from the mtDNA to determine your place in the genealogical tree that connects all humans through the direct female line. We classify your mtDNA into one of 27 basic branches (mitogroups) on this tree. However, the SNPs we test allow for a much finer classification into hundreds of different smaller branches. Where do you belong in the great human female line genealogy?

Apologies if I seem a bit grouchy but so far all these different tests have just been so confusing, when all I want to know is where my ancestors came from.

[cacio]

The starred groups are residual groups. In your case, this means that you belong to R, but not to any of the other branches shown in the graph. To make another example, I was assigned to U* by decodeme, but I know from other tests that I am U1a, a subbranch of U that decodeme does not test for.

There are very many branches in R, not shown in the tree. Most of these branches are in India, some also in SE Asia. There are also some related to H and V (eg R0, HV*) present in Europe; from the graph it's hard to see whether you belong to one of these.

So I can see different possibilities.
1) there was a no-read in the H SNP. This SNP should have shown that you were H, but there were no results there (have you checked 14767?)
2) decodeme doesn't test for groups like HV or R0 (which are mostly middle-eastern or mediterranean), and you belong to one of these
3) you belong to one of these other R* lineages.

Nothing is impossible, but 3) are extremely rare in Europe, that's why 1) and 2) seem more plausible.

As for decodeme vs other companies, remember that decodeme tests 1 million SNP throughout the DNA, this is a huge amount of information. The mtDNA is only a very tiny fraction of the DNA. Companies that tests the mtdna only are more precise or informative in that respect. But decodeme gives you all the rest.

cacio

[burto]

Hi Cacio,
How do I find the 14767 you mention? On the advanced bit the closest number I can see is 1438G?

[darroll]

Link to DeCodeMe's R* through their demo below looks completely seperate to H to me:
http://demo.decodeme.com/ancestry/female-line-info/R*
I can't even figure out what I thought is supposed to be the most reliable (MtDNA)
Taken from DeCodeMe:

Mitogroups in the deCODEme Complete Scan

The deCODEme Complete Scan uses 163 SNPs from the mtDNA to determine your place in the genealogical tree that connects all humans through the direct female line. We classify your mtDNA into one of 27 basic branches (mitogroups) on this tree. However, the SNPs we test allow for a much finer classification into hundreds of different smaller branches. Where do you belong in the great human female line genealogy?

Apologies if I seem a bit grouchy but so far all these different tests have just been so confusing, when all I want to know is where my ancestors came from.

Did you try NG?
They have the most information.

[burto]

Cacio,
Is the number you mention found on the advanced page of the maternal lineage page, if you scroll down through all the Mito numbers? If so what am I looking for? There are frequencies for each allele on that list also?
Darroll I've not done the NG test as I wasn't specifically interested in my mtdna. However given the strange South Asian matches I've had in autosomal tests has made me wonder if it's worth looking into. Did think all basic mtdna tests could at least tell you what mito group you're from, but it seems not.

[cacio]

Burto:

I don't think the genome browser works with the mtDNA (if someone knows how to, I'd like to know). So what you can do is the following. Go to advanced - data download -download your raw genotype data.

You'll get a huge text file with all your results for all chromosomes. The mtDNA is at the end of the file, So open the file with a text editor (I don't know if excel is powerful enough to read such a big file. Plain text editors or word should do). Go to the end of the file. You'll see that the last 100 lines or so all start with Mito. These are the results for the mtDNA.

The line I was talking about is the following:
MitoT14767C,C/T,M,14766,+,TT

This means that it's position 14766 in CRS. C and T are the two values that the marker can take. Haplogroups HV have CC, while everybody else has TT (I am U, so I have TT, as above). If you have CC, you should be in H. If you have TT, you should be outside H.

Another indicative line is
MitoG2708A,A/G,M,2706,+,GG

G indicates "outside H", A indicates H.

Of course, testing mistakes are always possible, as are "backmutations" (ie H people who for some reason mutate 14766 back to T).

You can peruse the rest of these mtdna results to see whether you have any particular mutation. What I was saying before, though, is that there are only about 10 position tested in HVR1 (that is, 16000 or above), so you could have other mutations that have not been tested.

cacio