Announcement

Collapse
No announcement yet.

Autosomal DNA: Are We Being Hoodwinked??

Collapse
This topic is closed.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Autosomal DNA: Are We Being Hoodwinked??

    OK, sorry for playing the devl's advocate with this... but I wish to pose a simple question to this forum's community:

    With the exception of adoptees and those who may not know their immediate ancestry, how do the cousin matching utilities offered by 23andMe, dna.ancestry, and FTDNA lead to any substantial finds of genealogical value? Afterall, it has been labelled as genetic genealogy.

    I contend that without existing papertrail supported family trees, matches cannot lead to determining the MRCA. And, with existing papertrails, there is virtually no way for the dna match to verify the papertrail... at least until the databases and tools become enormously more sophisticated than they are today. Thus, the playing field of genealogy remains in the same ballpark: 200 years with greatly diminishing returns therebefore.

    I would appreciate emotional responses by those heavily invested to be witheld. Let's talk about the science, the math, the probabilities, the RESULTs.

    Final note: hoodwinked may imply intent of action. I do not necessarily argue there is intent to deceive, as the proprietors of these businesses may truly believe this is going somewhere. But I am questioning how matching a 5cM segment with someone has to do with mainstream GENEALOGICAL RESEARCH. In the very least, are we being overly optimistic?

    wikipedia definition: Genealogy (from Greek: γενεά, genea, "generation"; and λόγος, logos, "knowledge") is the study of families and the tracing of their lineages and history.

    wikipedia definition: Genetic genealogy is the application of genetics to traditional genealogy.

  • #2
    I don't think anyone is being hoodwinked. The information is out there for those who seek it. For example see http://www.familytreedna.com/faq/answers.aspx?id=17#628 . This link clearly shows it's not an exact science.

    A 5cm match well is anything but a match. It's more likely to be identical by chance than a very distant ancestor. Determining common ancestry is NOT an exact science, and can clearly be determined given the information FTDNA or 23andme provide.

    Now do these companies clearly say in their advertising the gory details? Of course not, well except the CEO FTDNA did publicly state MtDNA testing is not very useful for genetic genealogy.

    Anytime you buy technology, and purchasing DNA testing is buying the latest technology, the customer really needs to understand to some degree of details before purchasing. It's no different than buying a car or latest electronic gadget from that respect.

    Comment


    • #3
      FF matches

      Sad to say, so far I have discovered nothing with my FF matches. No brick walls have fallen, no unknown cousins have been discovered. What few matches I can confirm all come from my maternal side. Either I got most of my DNA from my mothers side or none of my paternal relatives have tested. I don't think either of those two reasons are valid.

      I don't really know the science behind all the genetic testing, but in order for anything meaningful to happen, the science must be able to determine to some degree which side the match is coming from and what generation it is matching. Currently this is not possible and we will forever be uncertain about our proposed matches.

      Comment


      • #4
        Originally posted by sspeters41 View Post
        I don't really know the science behind all the genetic testing, but in order for anything meaningful to happen, the science must be able to determine to some degree which side the match is coming from and what generation it is matching. Currently this is not possible and we will forever be uncertain about our proposed matches.
        Well if you get parents tested you can run a phasing tool to determine which side mother or father a match exists. Currently 23andme offers this no additional charge. FTDNA has said they will provide this , but can be done with a little effort on your own. See http://www.math.mun.ca/~dapike/FF23utils

        Determining which generation really can not get a whole lot better. Science is limited due to the fact you only get 50% of each parents DNA and the rate that gets passed per generation can be so drastically different past the parent child generation. Going far back it becomes impossible to determine between a segment match due to common ancestor and just a match by chance.

        Comment


        • #5
          Nahhhhh!

          I think for people who come here totally uninformed probably think they are not getting what they paid for! I cannot believe people come on here or the other sites and won't share because haplogroups do not match. What...this is an autosomal test..that is testing DNA from both sides! If you hang on to that notion you ain't going find any cousins!!! There are also many who know about genetics and become disappointed that these tests are limited and not giving enough specifics for their liking. These tests are on a fast track but can't keep up with what we all want and expect them to give us.
          I was excited about the autosomal because it was going to "maybe" shed light on my unknown side...At least the ethnicity part anyway. Well it does seem that many of the populations are too close to call and they don't have all the populations tested to distinguish at least my one side..the unknown side! I too seem to be getting all my matches from my maternal side and no real clues to my fathers. I am sure it is his population are not testing. Until these populations are included I will have to be patient and happy with what they Have provided me. I have many Irish and Acadian matches...I am both It picked up my Spanish with a few matches here with my Ancestors correct names that's good! Lastly it tells me I am lots of Southern Europe with ME
          thrown in. Not so good as I don't have matches to prove it. All in all I have proven what I already knew and have gotten clues to what I did not know. I have learned so much more along the way...yes it is worth it!

          Comment


          • #6
            Originally posted by thetick View Post
            Well if you get parents tested you can run a phasing tool to determine which side mother or father a match exists. Currently 23andme offers this no additional charge. FTDNA has said they will provide this ...
            Actually FTDNA is way ahead of 23andMe on this matter because 23andMe's "phasing" is merely a list comparison between a child and Parent(s). not true phasing.

            At Ftdna you can perform the equivalent by using "in common with" or "not in common with" using a parent. The fact that one has the option of using ICW or NICW with anyone on your match list, not just parents, makes FTDNA's offering much more powerful and useful

            Comment


            • #7
              I have had a few successes; but my methodology is a bit different than others. I have tested a wide field of closely related relatives of mine to maximize the diversity of the genome:

              1. It allows me to "see" a bit further than if I had just tested myself. Each relative is like a radio telescope listening for the signal; and I have a large array of them.

              2. It allows me to narrow down to the quarter or eighth of my family that the match is coming from. The hunt becomes easier.

              I haven't discovered new ancestors, yet, but have assembled several good clues about where to look. I have confirmed several lineages of mine that were "iffy"; there was a weak link in the generational chain.

              I think that as more people get tested & more people upload gedcoms, the chances of finding the common ancestor will become a bit easier.

              Having tested a series of my parents' first cousins & second cousins, I can say that the technology works. All cousins share centimorgans that were within range of what is expected: 838 cms for first cousins & 209.5 cms for second cousins. Some are somewhat higher, as expected & some are somewhat lower, as expected. The fact that Family Finder was able to find & confirm these kinships speaks for itself.

              Timothy Peterman

              Comment


              • #8
                Optimistic

                I always see the glass full, just my nature. I understand how people can get disappointed with their dna testing results, but you have to persevere, and communicate with project admin folks who can guide you. I am an adoptee, and I have been fairly lucky so far. At this point in time, I know what specific family line I come from as far as my paternal side, just nailing down the actual branch, which will take some more time. I have just ordered the Gene 2.0 test, and I think that will give me some more details about my lineages. If you want to find out more about your lineages, you need to communicate with project admin folks, who have been invaluable to me, guiding me in what snp(s) to test for, etc. Always look at the glass as half full.

                Comment


                • #9
                  I do believe there is value in autsomal testing although I don't yet have hard proof. I do have some examples:

                  1. My sister is searching for her biological father's name. She has been lucky enough to have matches that point to a specific couple as her likely ancestors on her paternal side. We have also identified a potential paternal candidate who shares this couple as his ancestors. Unfortunately he has passed away, and we are currently trying to find a sibling or other close relative that is willing to test our theory.

                  2. I have managed to find a common ancestor with several matches. In these cases, I realize it's still not "proof" that the shared dna is actually from that common ancestor. However, I think with time, we will develop better tools that will help confirm this matches and their corresponding paper trails. In all of these cases, I already had the common ancestor in my gedcom, so these cases don't really contribute to my genealogy as of yet.

                  3. I have several lines with brickwalls. One example is my McCall line. I have tried for years to identify my James McCall's parents. I now have a few matches that descend from a specific McCall line out of Pennsylvania. With enough dna evidence, I can then focus my traditional research on that particular McCall line, rather than trying to investigate all of the McCall lines that were in the US in the late 1700s. When will I consider that I have enough dna evidence? I'm not sure, and I'm not there yet. However, as more people test, and as I test more people in my family, patterns begin to develop. I firmly believe there will be a snowball effect, where the patterns develop to a point that much of the work begins to pay off. Maybe I'm overly optimistic, but I am having fun with it.

                  I hope others can post more specific results. I think your post is timely. I think that most newbies are initially extremely excited (as I was) when first receiving their results. As time goes by, many people become discouraged about their apparent lack of success. Those who stick with it through this phase, I think will eventually see the pay off. I certainly thought this process would be much easier, but I'm not too upset that it isn't. As I said, I'm having fun. It's like my early days of genealogy research.

                  Emy

                  Comment


                  • #10
                    To determine which side of your family a particular genetic match falls on you need to have one parent or some who is related to one parent tested. In my case, I have a half-brother. If someone matches us both, then that person is probably a paternal match. If someone matches only me, then that person is probably a maternal match. If someone matches only my brother, then that person is probably a maternal match for him.

                    Having 800 and growing matches at 23andMe and about 50 matches at FT-DNA, I disagree with the contention that FT-DNA is more useful.

                    Comment


                    • #11
                      Originally posted by Bertp View Post
                      Actually FTDNA is way ahead of 23andMe on this matter because 23andMe's "phasing" is merely a list comparison between a child and Parent(s). not true phasing.
                      I don't really agree with this statement since 23andme has Ancestry Composition that is phased with one or both parents. FTDNA's Population Finder is not even in the ballpark with 23andme's AC with no phasing and much poorer results.

                      Comment


                      • #12
                        I see one of the primary values of DNA testing is to verify that one's pedigree is correct. Now some may argue that finding a DNA match that shares a documented common ancestor is not 100% proof that the two lines are in fact correct, but it probably is for any reasonable person. When I get a DNA distant cousin match, and determine we both have lines going back to the same obscure tiny village in Europe, that is rather compelling evidence. The odds of that happening by chance are so remote that ii is not really rational to worry about it. This means of verification is especially important if one knows there is a shaky link in one's pedigree, such as two emigrants names Ole Jonsson who were both born in 1840 and both emigrated to Minnesota in 1865, but came from different places. In the past, one just had to hope you made the right guess as to which one to trace. Now if one finds a DNA matching distant cousin descended from the grandfather of the Ole you picked, you can be pretty certain you chose the correct one. If there are no such matches, you know you should perhaps look at the other Ole!
                        Also, finding distant cousins sometimes brings one family stories and relics that passed down their line, but not mine. The test results can also lend credence to family legends. In one of my lines, the undocumented oral tradition was that some ancestors had migrated from what is now north central Sweden to northern Norway in the 1600s. This was too early to have much if any paper documentation, as moving records were still a hundred years away. However, as I am finding a number of distant cousin matches with roots in north central Sweden (and none of my other lines are close to that area), I am confident in now concluding the oral tradition is probably factual rather than mythical.

                        Comment


                        • #13
                          Autosomal DNA, provides less and less certainty as you move back further in time. This makes sense given how the process of recombination works. As other have noted, this doesn't mean it is useless even when we're talking about small or relatively uncertain matches. You just need to know what the resolution and probabilities are and be patient. The picture may be fuzzy but odds are good it will continue to improve as more people test.

                          You can see the effects of "noise" in autosomal DNA even in fairly close matches.

                          The second cousin of my Mom that I recently discovered via a match on 23andMe is a good example. He doesn't match my Mom at all on the X chromosome yet he and I share a small match. That small match is likely the result of my Mom's two X chromosome's recombining before being passed along to me.

                          He and I also have a longer match on chromosome three than he shares with my Mom. Again, this is noise. My paternal side recombined with my maternal at that spot and by random chance I ended up with matching segment that is longer than my Mom.

                          To call those examples noise is a bit misleading though. In all likely hood they represent clues that our newly discovered relative shares a similar ethnic background with us and the 23andMe Ancestry Composition tool bears this out.

                          Reading autosomal DNA is not like reading tea leaves but most of the matches we're going to get here and on sites like ancestry or 23andMe are very distant and thus have low information content, particularly if you look at each one in isolation. Based on my limited understanding of the science I don't see that changing.

                          That doesn't mean more information can't be found as tools are developed to look at all of ones matches. It should become easier to separate signal from noise as the amount of information available increases. I suspect this is particularly true if you have the ability to test close relatives.

                          In the meantime it's going to take a bit of luck and a lot of detective work to extract all the information that is available to us.
                          Last edited by 1_mke; 5 March 2013, 10:12 AM.

                          Comment


                          • #14
                            Family Finder has afforded me much despite the recent issues... I don't have the luxury of phasing because my mother is very ill and my father is one of those who will never test. I have not tested elsewhere. I am content with FDNA, Gedmatch and my extensive tree on Ancestry.com I am the only one in my family who is engaged in genetic genealogy and many family stories have been confirmed. And to my surprise I discovered that I am a 8th and 10th cousin to my husband! A small handful of his Ashkenazi matches are mine as well and for a man who was adopted this indeed is a treasure.

                            I am able to tell with great accuracy which matches fall where, as in which are paternal and which are maternal. My father's family is Sephardi, up until 1840 on his paternal line they were in Mexico, since the 1600's, before that Spain, before that Italy, the Middle East etc. His mother crossed the border in 1942.

                            My maternal line is Ashkenazi with a small amount of Scot/Irish/British thrown in. My maternal - paternal line is German/Russian/Lithuania/Ukrainian Jewish. My mother's direct maternal line is a mix of Ashkenazi with the Scot/Irish.

                            I have 130 matches now, lost 4 pages. Of those - 33 are paternal links of the 33 I have confirmed a 10th cousin match and through that match all of my other paternal matches match him. The other 97 matches reflect my maternal line wonderfully. I have 2 who provide the Finnish connection, 5 who are Cohen and so forth.

                            For the most part I am very happy with Family Finder however like many I am not thrilled with Population Finder. Thanks to Gedmatch I've been able to locate even more Ashkenazi matches and by more I mean over a 100.

                            Not all of us will have successes but many of us do. Yes, things could be better however I try and remind myself that the science is still evolving, that there will be growing pains and despite everything it is exciting to be involved with it as it matures.

                            Comment


                            • #15
                              I think that Family Finder really should be promoted & marketed as a family test, rather than as an individual test. A good family finder project will contain several individuals (siblings, their cousins & 2nd cousins).

                              This could be done by offering the first test for $289, the next five for say $199 each, the next 10 for say $179 each, and so forth. In a way, they have done this, by having periodic sales. Those of us who test multiple relatives have figured out when the sales are.

                              Timothy Peterman

                              Comment

                              Working...
                              X