Announcement

Collapse
No announcement yet.

Confused and Disappointed

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Confused and Disappointed

    Just revisited the Genographic site. My wife went from U to F which made sense since not alot of Finns passed thru India on their journey up north. I went from R1a to R1a1.

    How do these changes i groups occur?

    Secondly, ran my wife thru MitoSearch, and not a single match. Is the database that small, or are mtdna Haplogroup F folks just "special".

    Best, Glenn

  • #2
    Glenn,

    A major change from U to F sounds like something caused FTDNA or the lab to question your wife's original results and doublecheck them. You should definitely contact them to find out what happened and confirm that the current results are really the correct results. You can email [email protected] and give them your wife's Genographic Project ID number (don't post it here).

    For your own results changing from R1a to R1a1, I believe they just updated the database to rename the haplogroup subclade that goes with your SNP result. The entire haplogroup R tree can be found here: http://www.isogg.org/tree/ISOGG_HapgrpR.html. You'll see that R1a1 is defined by a SNP at M17 and M198 and is downstream from R1a. These types of refinements or nomenclature changes happen periodically as the science advances and new information becomes available.

    Have you transferred both your and your wife's results to FTDNA? If not, you definitely should, as you'll then benefit from the matching that they automatically do between your DNA and everyone else in their database, plus you'll be able to order advanced tests if you're interested. To transfer your results to FTDNA, log in to your Genographic pages and at the very bottom of the page, you'll see a link to 'learn more' about Family Tree DNA. Just follow the instructions there and you'll be walked through creating an FTDNA account. Once you have your FTDNA account, log in and you can explore the various tabs, including the Y-DNA Matches tab (on yours) and the mtDNA Matches tab (on your wife's). You'll have to create the two accounts separately -- there's no way to combine them.

    Mitosearch and Ysearch don't have nearly the amount of results as FTDNA's database since those databases are optional and not everyone chooses to upload their results to them. However, they are open to everyone who tested with any company, so definitely search there as you may also find matches who didn't test with FTDNA and therefore wouldn't show up in your list of matches at FTDNA.

    Hope this helps!

    Comment


    • #3
      Glenn:

      I know of one case of a person who was originally assigned to mtDNA U1a, but later moved to F. It was a pretty special case, though. Presumably, your wife has a signature typical both of some U sub-branch and of F, and for some reason she was reassigned. As efgen said, you can send an email and ask them why they reassigned. The difference between F and U comes from the coding region outside HVR, so it may be interesting to see if they actually tested that. If you want, you can post her mutations here, there are some knowledgeable people who have access to large databases and can verify the odds ratios. Obviously, the genographic project has very few people from Asia, so that's why there are no matches.

      Note that U is more common in India than F (>10% versus 2%). F is more typical of East Asia. U5 is typical European, but other U subgroups are present in India in good numbers.

      cacio

      Comment


      • #4
        Originally posted by rojamg
        Secondly, ran my wife thru MitoSearch, and not a single match. Is the database that small, or are mtdna Haplogroup F folks just "special".
        Glenn, what's her sequence?

        Comment


        • #5
          Wife values

          Here's my wife's stuff: Was U, now F
          16188T, 16266T, 16304C, 16311C, 16356C, 16456A, 16519C

          Thanks for your suggestions, Glenn

          Comment


          • #6
            rojamg:

            I quickly browsed a couple of papers on India and found little of any help. The closest thing is a 266 304 311 355 356, which sounds similar, in
            http://jorde-lab.genetics.utah.edu/e...20lineages.%22
            mysteriously classified as R1, not F. I believe F is not a subgroup of R1, but I really know nothing about Asian mtdna groups. An often quoted article by Palanichamy doesn't have any R1, R9 or F to show. Let's see what vraatyah says.

            Of course, since India is very diverse and to some extent not well studied, it is possible that your sequence doesn't fit any well established European/East Asian haplogroup, but represents rather an unknown ancient Indian lineage (in any case, way more ancient than those R1a Y-chromosome newcomers. By the way, despite the fact that I am Italian, my Y-haplogroup L is fairly Brahmin too.)

            cacio

            Comment


            • #7
              Originally posted by rojamg
              Here's my wife's stuff: Was U, now F
              16188T, 16266T, 16304C, 16311C, 16356C, 16456A, 16519C
              F and U are both mtDNA subgroups of R:

              http://en.wikipedia.org/wiki/Haplogroup_F_(mtDNA)

              Comment


              • #8
                Originally posted by rojamg
                Here's my wife's stuff: Was U, now F
                16188T, 16266T, 16304C, 16311C, 16356C, 16456A, 16519C

                Thanks for your suggestions, Glenn

                As Casio pointed, it's R5, an independent Indian subbranch of R, not F (F appeared to be common in South India too, ultimately it has more Eastern origin). For the complete R5 sequences you may inspect Palanichamy et al. 2004 paper, though it does not provide any exact matches. The closest variant I know is one from a paper by Barnabas, 2005 y:

                188-223-266-304-311-325-356
                (between 16024 and 16382, so it may have the same state at 16456 site)
                -7853HincII -10394DdeI -10397AluI +13633HaeIII

                the latter marker indicates being in R5


                I suspect either your wife's sequence or one of Barnabas contain errors: without having 325 it's unlikely to get 188 because many R5s have 325 and only two have 188. If these 2 variants are related, the sequences with 188 is a further branch from 325 node. So, your variant lost 325, or it was overlooked during the sequencing, otherwise your variant and one from Barnabas' paper got the transition at 188 independently.


                Valery

                Comment


                • #9
                  Glenn, you may ask you lab to check the traces for 16325. Perhaps they indeed overlooked this mutations. I think, it's for free, just to open a file.

                  Comment


                  • #10
                    Reply from FTDNA

                    Response from E-Mail:

                    ""There are two systems for determining a haplogroup; one based on the HVR1
                    results and the other based on additional haplogroup testing. Sometimes
                    these will be in conflict. In your wife's case, her HVR1 region looks like
                    she ought to belong to haplogroup U, but her haplogroup test clearly
                    indicated that she is not in U. This is a good example of why it's
                    important to test for haplogroups rather than depending on the HVR1 region,
                    since the latter can sometimes be misleading.

                    A clerical error when the results were initially uploaded caused the HVR1
                    haplogroup determination to be used rather than the haplogroup test result.
                    Our mtDNA team periodically reviews the database of the thousands of tests
                    performed to check for any errors. They caught this one, and had us correct
                    the haplogroup assignment. Based on a combination of the haplogroup test
                    results and the HVR1 region, her correct haplogroup assignment is F.""

                    I still waiting from my beginers guide to DNA testing from amazon.com May be then the fog of DNA terminology will lift. I assume the HVR1 region is a specific region of genetic code whose variations are the basis for genetic analysis. What exactly is a a Haplogroup test if not the summation of variation of a specific DNA strand?

                    I'll forward the request for additional tesing. I'd hate for my wife to live in U, F, R purgatory.

                    Bests, Glenn (R1A1)

                    Comment


                    • #11
                      Originally posted by rojamg
                      Response from E-Mail:

                      ""There are two systems for determining a haplogroup; one based on the HVR1
                      results and the other based on additional haplogroup testing. Sometimes
                      these will be in conflict. In your wife's case, her HVR1 region looks like
                      she ought to belong to haplogroup U, but her haplogroup test clearly
                      indicated that she is not in U. This is a good example of why it's
                      important to test for haplogroups rather than depending on the HVR1 region,
                      since the latter can sometimes be misleading.

                      A clerical error when the results were initially uploaded caused the HVR1
                      haplogroup determination to be used rather than the haplogroup test result.
                      Our mtDNA team periodically reviews the database of the thousands of tests
                      performed to check for any errors. They caught this one, and had us correct
                      the haplogroup assignment. Based on a combination of the haplogroup test
                      results and the HVR1 region, her correct haplogroup assignment is F.""

                      I still waiting from my beginers guide to DNA testing from amazon.com May be then the fog of DNA terminology will lift. I assume the HVR1 region is a specific region of genetic code whose variations are the basis for genetic analysis. What exactly is a a Haplogroup test if not the summation of variation of a specific DNA strand?

                      I'll forward the request for additional tesing. I'd hate for my wife to live in U, F, R purgatory.

                      Bests, Glenn (R1A1)

                      their comment does not seem to be of great importance. At least, until they list the markers included in multiplex.

                      Comment


                      • #12
                        rojamg:

                        as vraatyah explained, the initial answers from ftdna are just standard clerical responses of little content. Only after some insistence and more specific questions does one reach more knowledgeable people in the company. For instance, in your case, what they ought to have said was exactly what they tested. They say they didn't find the marker for U. But I don't think they tested for F, they just guessed it based on 16304, which is also present in F.

                        Regarding your question about HVR1, you are right that it is a specific region. Mtdna is composed of around 16600 letters, HVR only comprises the last 600. However, it is often not enough to determine the haplogroup. The mutations defining the haplogroup are usually in the other parts of the mtDNA. Moreover, unlike for the Y chromosome, mutations appear independently in more than one haplogroup, so often one has to check a whole set of them, not just an isolated one. As vraatyah said, I don't think FTDNA at this point has the test for the mutations defining most non-European haplogroups, I think they do have that for haplogroups like H and U, which are common in Europe.

                        What FTDNA does usually is to look at HVR1, and then _predict_ the haplogroup based on their database. Their system works relatively well for European haplogroups, since they have a huge database. But it breaks down for the non-European ones. And this also explain why they tend to overpredict European haplogroups (you are not the only case of such misassignment) - customers are usually of European origin, so H and U are usually good guesses.

                        cacio

                        Comment


                        • #13
                          I have to disagree with cacio's assessment of the email from FTDNA. The way I read it, they predicted the haplogroup from HVR1 as U, but their haplogroup test showed it to be F. However, through a clerical error, they uploaded the predicted haplogroup instead of the tested haplogroup to the GP database. They then caught their own error while doing a review, hence the change.

                          I believe FTDNA always does a haplogroup test for mtDNA since many haplogroups are defined by coding region mutations rather than HVR1/2 mutations.

                          Glenn - For many mtDNA haplogroups, the defining mutations are not in the HVR1 or HVR2 regions that most of us have tested. Instead, they are in the coding regions, so FTDNA has to test for additional mutations in the coding region in order to correctly identify the haplogroup.

                          vraatyah - FTDNA doesn't automatically provide the list of coding region mutations that they test to determine haplogroup. I don't know if they would provide the info if asked.

                          Comment


                          • #14
                            efgen:

                            of course, they may have tested for F. I am skeptical, however, because I have seen previous cases of misassignment and seen similar emails before for more than one case. As you suggest, the only way to know is to ask. And to ask precise questions. Usually, the first answer one gets is from a customer rep, a non-scientist, who only answers with standard sentences that don't mean anything (everything is always a clerical error), as in this case. One gets through and talks to a scientist only after a couple of emails with a very precise question such as: is the assignment based only on HVR1 or were other locations in the coding region tested, and if so which ones? Or an even better question, this sequence looks similar to (cite academic article): how is your assignment reconciled with this?

                            FTDNA does not usually test for coding region mutations, they may do only in certain dubious cases and for basic European groups like H and U. And, to be fair, they cannot do otherwise. There are literally hundreds of haplogroups when moving out of Europe, essentially they'd have to test the whole mtdna - not something they can do for $100. Which I suppose is the reason why we don't have a mtdna SNP haplogroup test being offered by companies (while there are more than a few companies offering Y SNP tests.)


                            cacio

                            Comment


                            • #15
                              Originally posted by cacio
                              FTDNA does not usually test for coding region mutations, they may do only in certain dubious cases and for basic European groups like H and U. And, to be fair, they cannot do otherwise. There are literally hundreds of haplogroups when moving out of Europe, essentially they'd have to test the whole mtdna - not something they can do for $100.
                              There are not hundreds of haplogroups -- maybe hundreds of subclades. And FTDNA most certainly does determine the haplogroup by testing the coding region -- finally found the reference, in FTDNA's new mtDNA Tutorial:

                              http://www.familytreedna.com/mtDNA_tutorial.html#9

                              Subclades are another story. The only mtDNA haplogroup that FTDNA currently offers a subclade test for is H. For all other haplogroups, the only way to test for the subclade is to order the mtDNA Full Sequence test.

                              Comment

                              Working...
                              X