Announcement

Collapse
No announcement yet.

Best Place to Begin?

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

  • Best Place to Begin?

    Since people on this forum have had success with DNA, here it a dilemma I have that I could use some advice on. Where does someone begin when their family tree is wrong?

    It started with weird ethnicity results that made absolutely no sense with what I had in my tree. So, I began contacting people within my tree (people within the 3rd cousin-ish range) who were supposedly related to me and asking them to upload to Gedmatch to see if we were indeed related, but haven't been able to find a connection. What now? Obviously something is wrong, and right now I don't have the option of testing more family members. I haven't had the best luck contacting people either, most people haven't responded or they themselves are in a predicament where they know nothing about their family either.

  • #2
    First of all, if the only ethnicity results you have are from FTDNA, ignore them. If you have tested at all 3 companies and they all say the same thing, you might want to pay some attention. But FTDNA's are the worst.

    Where did you get your tree? Was it through years of careful research with primary documents as sources? Or did you just copy someone's online tree? How do you know it's wrong? If it's just copied it probably is wrong, unless you had the good fortune to copy the tree of a meticulous researcher.

    I haven't had much luck getting replies here. And I don't have any idea what common ancestors I share with my 2nd to 4th cousins here, much less 3rd to 5th. If I hadn't tested at Ancestry too, I'd be thinking maybe my tree was wrong too. But at Ancestry, my mother has plenty of very close matches to verify her Calvin line, plus plenty of very close matches to verify two of her German lines. I haven't gotten any matches to verify any of the ancestry of her French grandfather who immigrated in 1905. Less than a half dozen matches here in these lines and they are quite distant.

    As far her brick wall lines go, I'm not getting anywhere. It's hard to find something when you don't know what it is you're looking for.

    Comment


    • #3
      Originally posted by MoberlyDrake View Post
      First of all, if the only ethnicity results you have are from FTDNA, ignore them. If you have tested at all 3 companies and they all say the same thing, you might want to pay some attention. But FTDNA's are the worst.

      Where did you get your tree? Was it through years of careful research with primary documents as sources? Or did you just copy someone's online tree? How do you know it's wrong? If it's just copied it probably is wrong, unless you had the good fortune to copy the tree of a meticulous researcher.

      I haven't had much luck getting replies here. And I don't have any idea what common ancestors I share with my 2nd to 4th cousins here, much less 3rd to 5th. If I hadn't tested at Ancestry too, I'd be thinking maybe my tree was wrong too. But at Ancestry, my mother has plenty of very close matches to verify her Calvin line, plus plenty of very close matches to verify two of her German lines. I haven't gotten any matches to verify any of the ancestry of her French grandfather who immigrated in 1905. Less than a half dozen matches here in these lines and they are quite distant.

      As far her brick wall lines go, I'm not getting anywhere. It's hard to find something when you don't know what it is you're looking for.

      I'm not just going by ethnicity, not by a long shot. I know that is faulty methodology. For what it's worth however, I have uploaded to multiple sites and they have all pretty much said the same thing. The percentages of some of the ethnicities (very large) plus the fact the tester is being matched with actual people of that ethnicity leave little doubt that it is for real.


      The tree I have constructed I have been working on for a couple years & is well researched and documented with original sources (when I first started genealogy I made the mistake of just copying information, but have since learned better!). I know my tree is wrong because I'm not matching with people I'm supposed to be matching with, and the ethnic percentages make absolute zero sense. I had connected with some relatives (more distant) prior to the DNA test and we had already figured out our relationship/ common ancestor. Yet, no dice on a genetic match.

      Comment


      • #4
        Originally posted by lab View Post
        I had connected with some relatives (more distant) prior to the DNA test and we had already figured out our relationship/ common ancestor. Yet, no dice on a genetic match.
        How distant? I suppose they are close enough that you have a fairly good probability that they should match on a DNA test?

        Do you know which branch of your tree the error is likely to be in?

        I had to cut out a whole branch of my tree when Y-DNA testing revealed an NPE. It took me a year to sort out where the NPE occurred and prove it by DNA testing. That gave me the name of my mother's great-grandfather. I easily found his parents, but I've been trying for years now to find their parents. When you can't find documents, it appears atDNA testing is not very helpful at the 3rd great-grandparent level.

        Is the error in a line where you might be able to use Y-DNA testing if you could locate some willing male cousins?

        Do you have matches closer than 3rd to 5th cousins? I'd work with the closest.

        Comment


        • #5
          It really depends on your goals. If you live in America, are you wanting to see where you ancestors came from or just delve into your ethnicity?

          Start with your family tree and use autosomal DNA and related testing to help fill in the gaps and get more answers.

          I do think ancestry.com probably has the most complete family tree service. Start by finding out the full names of your parent's parents. Once you have that information, start looking at the census records on Ancestry.com. It's very easy in most cases keep finding grandmothers/grandfathers at that point. Once you hit the 1700s it will get a little tricky and that's where DNA evidence helps.

          It's also helpful to find out where your great parents settled and will point to clues where they may have come from.

          Comment


          • #6
            Originally posted by MoberlyDrake View Post
            How distant? I suppose they are close enough that you have a fairly good probability that they should match on a DNA test?

            Do you know which branch of your tree the error is likely to be in?

            I had to cut out a whole branch of my tree when Y-DNA testing revealed an NPE. It took me a year to sort out where the NPE occurred and prove it by DNA testing. That gave me the name of my mother's great-grandfather. I easily found his parents, but I've been trying for years now to find their parents. When you can't find documents, it appears atDNA testing is not very helpful at the 3rd great-grandparent level.

            Is the error in a line where you might be able to use Y-DNA testing if you could locate some willing male cousins?

            Do you have matches closer than 3rd to 5th cousins? I'd work with the closest.

            How did you go about finding the NPE? That's probably what I'm dealing with here but I don't really know where to begin since I can't really test any other family right now. I have a couple 2-4th cousins but not many and none have responded to me.

            Comment


            • #7
              I had my 1st cousin do a Y-DNA test because I had a brick wall in my Mobley line and thought it might help. I suppose you know that a Y-DNA test follows the father's father's father's etc. line. When his results came back, he matched 5 or 6 Hamptons (more now, 7 years later) and absolutely no Mobleys. Checking censuses revealed that there were Hamptons living not far from the Mobleys in several generations and locations, but I decided to start with a situation where a Hampton was the next household to my mother's supposed great-grandfather, because of what I thought were suspicious circumstances surrounding her grandfather's placement in an orphanage. I had already done a Family Finder test on my mother. I managed to locate a descendant of the brother of the Hampton suspect to do the Family Finder test and when his results came back he and my mother were 3rd cousin matches.

              This was back in 2010-2011 when atDNA testing was new and Ancestry hadn't yet gotten into the act. Now my mother has a few more very close matches descended from the Hampton neighbor. Records quickly revealed his parents names. But that just gave me two brick walls, which neither traditional research nor DNA have broken through.

              Comment


              • #8
                lab,

                You haven't said what ethnicity results you are getting and what you were expecting. Are you certain there's no logical explanation for the difference? What information are you basing your expectations on?

                You mentioned that you don't match people from your tree "within the 3rd cousin-ish range". What relationships exactly? As far as I know (and this is being actively discussed in some groups) there are no proven cases of second cousins once removed (or closer) not sharing autosomal DNA, but it's widely accepted that 5-10% of 3rd cousins don't share DNA and around 50% of fourth cousins don't share DNA. So, unless you've tested people who are 2nd-once-removed or closer you can't be sure that the tree is wrong based on the lack of a DNA match.

                You said that you do match people who have the ethnicity you weren't expecting to see in your ancestry. How close are these matches?

                Comment


                • #9
                  Originally posted by DaveInGreece View Post
                  lab,

                  You haven't said what ethnicity results you are getting and what you were expecting. Are you certain there's no logical explanation for the difference? What information are you basing your expectations on?

                  You mentioned that you don't match people from your tree "within the 3rd cousin-ish range". What relationships exactly? As far as I know (and this is being actively discussed in some groups) there are no proven cases of second cousins once removed (or closer) not sharing autosomal DNA, but it's widely accepted that 5-10% of 3rd cousins don't share DNA and around 50% of fourth cousins don't share DNA. So, unless you've tested people who are 2nd-once-removed or closer you can't be sure that the tree is wrong based on the lack of a DNA match.

                  You said that you do match people who have the ethnicity you weren't expecting to see in your ancestry. How close are these matches?
                  What are the odds of not matching multiple 3rd cousins though? I have multiple people I connected with on my tree (before testing) that aren't matching, which seems odd to me but based off what you said could be possible.

                  I was expecting based off my research around 20% French, and 70-75% english/ Irish, the rest just miscellaneous. This tester came up as more than 50% of an ethnicity that hasn't in any way been accounted for (middle eastern). We are not Italian based off my tree, nor are they matching any Italians, which would be the only semi-plausible explanation. Instead, they are matching individuals from the Middle East. Some of the matches are distant but some are as close as 2nd cousin once removed range. I'm not sure how to make those things add up.
                  Last edited by lab; 6 May 2017, 05:07 PM.

                  Comment


                  • #10
                    I haven't had any unusual ethnicity results (everyone comes out mostly European with mostly British Isles or Western and Central Europe) that don't contradict the belief that the testers are pretty much from the British Isles.

                    However, trying to work out matches has been very difficult for one kit in particular. This was supposed to be the "fun & easy" kit, because (unlike some other kits I manage) all the grandparents are known and some lines go back much further.

                    Between FtDNA and Gedmatch there are only 5 recognisable matches. A 3C (via maternal grandmother), two 3C1R (one the son of the 3C and the other via the maternal grandfather), a 4C2R via the paternal grandmother and a very weak match on Gedmatch (not a match here) with a 1/2 3C via the paternal grandfather. That match is very weak so I am suspicious.

                    I have no idea whatsoever who everyone else is!

                    The odds of not matching multiple third cousins are slim. I have seen it on Gedmatch where two 3rd cousins match, but the 1st cousin of one of them doesn't. He does match other 3rd cousins though.
                    Last edited by ltd-jean-pull; 6 May 2017, 05:35 PM.

                    Comment


                    • #11
                      Originally posted by lab View Post
                      What are the odds of not matching multiple 3rd cousins though? I have multiple people I connected with on my tree (before testing) that aren't matching, which seems odd to me but based off what you said could be possible.
                      Do all of these 3rd cousins share the same parents?

                      Comment


                      • #12
                        Originally posted by MoberlyDrake View Post
                        Do all of these 3rd cousins share the same parents?
                        No, we all share the same great great grandparent, but through different children

                        Comment


                        • #13
                          One side of my family is extensively tested (26 cousins and growing). We had a case where cousin Simon got his results and his only match was around 9cM with one other cousin, which is such a small amount that it was inconclusive, given that Simon and the other cousin could easily have been distant relatives via some other line of undiscovered shared ancestry. We strongly suspected a "non paternal event".

                          However, about two weeks ago we got results for cousin Dennis. His match to Simon was above average for the relationship, and he also matched almost all other cousins at or above the expected range. Looking at Dennis's matches in the chromosome browser on Gedmatch you can see that Simon seems to have "freak" results (or the rest of us do, except Dennis) and seems to have inherited "grandma" wherever we got "grandpa" and vice versa.

                          I can really see this sort of effect looking at my nephew's results on his late mother's side. You'd expect his maternal DNA to be around 25% from each of his maternal great-grandparents, but he actually has a strong bias to his grandfather's side and the results vary from barely 19% from his grandmother's father to over 33% from his grandfather's mother (and that figure doesn't even include his X chromosome which he happens to have inherited entirely from his grandfather's mother). If you imagine that his first cousin (if he had one) might have a similar bias in the opposite direction, they would have a much smaller DNA match than average. Carrying on for two generations (third cousin level) there's a greatly increased chance that the cousins will share no DNA. The estimate that 5-10% of third cousins share no DNA (in significant segments) is based on the averages if you sampled 100s of cousins in 100s of different families, but the figures can be very different within individual families because of the potential for 1st cousins to have a low match with each other which affects the chances of their descendants having a match.

                          =======
                          As for the ethnicity... I don't know how far back your tree goes so it's difficult to comment much. But you were expecting French and have Middle Eastern instead? France had an extensive presence in the Middle East and Arab areas of North Africa, so Middle Eastern DNA hidden behind a French surname wouldn't be a major surprise.

                          Comment


                          • #14
                            Originally posted by DaveInGreece View Post
                            One side of my family is extensively tested (26 cousins and growing). We had a case where cousin Simon got his results and his only match was around 9cM with one other cousin, which is such a small amount that it was inconclusive, given that Simon and the other cousin could easily have been distant relatives via some other line of undiscovered shared ancestry. We strongly suspected a "non paternal event".

                            However, about two weeks ago we got results for cousin Dennis. His match to Simon was above average for the relationship, and he also matched almost all other cousins at or above the expected range. Looking at Dennis's matches in the chromosome browser on Gedmatch you can see that Simon seems to have "freak" results (or the rest of us do, except Dennis) and seems to have inherited "grandma" wherever we got "grandpa" and vice versa.

                            I can really see this sort of effect looking at my nephew's results on his late mother's side. You'd expect his maternal DNA to be around 25% from each of his maternal great-grandparents, but he actually has a strong bias to his grandfather's side and the results vary from barely 19% from his grandmother's father to over 33% from his grandfather's mother (and that figure doesn't even include his X chromosome which he happens to have inherited entirely from his grandfather's mother). If you imagine that his first cousin (if he had one) might have a similar bias in the opposite direction, they would have a much smaller DNA match than average. Carrying on for two generations (third cousin level) there's a greatly increased chance that the cousins will share no DNA. The estimate that 5-10% of third cousins share no DNA (in significant segments) is based on the averages if you sampled 100s of cousins in 100s of different families, but the figures can be very different within individual families because of the potential for 1st cousins to have a low match with each other which affects the chances of their descendants having a match.

                            =======
                            As for the ethnicity... I don't know how far back your tree goes so it's difficult to comment much. But you were expecting French and have Middle Eastern instead? France had an extensive presence in the Middle East and Arab areas of North Africa, so Middle Eastern DNA hidden behind a French surname wouldn't be a major surprise.
                            I've traced back to the late 1700's for most lines. My family has been in America (with the exception of one documented Irish GG-grandfather from Ireland in the mid-1800's) since the late 1700's/ very early 1800's. I don't know if this matters but the French in my ancestry isn't just from one person, it is from multiple people in multiple lines. Any full French ancestors were much farther back than I've traced.

                            To clarify, I was expecting French and do get approximately ~20% in most calculators, so that isn't actually the problem! The issue is I only get about 30% Northwest Europe on most calculators when I should be getting over 70%.

                            Comment


                            • #15
                              One thing you will need to remember when it comes to cousin matches is that though you match on paper, it doesn't mean that you will match through your DNA.

                              If you are looking at matches who are at the 3rd cousin on up level this is especially true. My brother and I have different match lists after a certain point. He has 3,406 matches to my 3,127. Of his 3,406, I match only 1,537 and he matches 1,536 of mine. But his cousins would all be mine on paper.

                              One of his cousins, who doesn't match with me, for example, matches him with 92cM and she also has done her Ancestry testing. At Ancestry we share a DNA circle for one of our ancestors, but we aren't sharing enough DNA to be called an actual match, though we have a paper match. Perhaps if we took our DNA to GEDmatch and lowered the amount of DNA to search for, a match might come up, or it might not.

                              And as to the amount a person can expect to have showing on the ethnic results, that too, is something that can't be called exactly due to the way DNA is inherited. The only time you get an exact amount of DNA from someone is when you get your 50% from each parent. Your parents don't hand on an exact 25% from each grandparent, and 12.5% from each great grandparent. My brother and I also don't share an exact ethnic estimate either. Don't worry too much about the ethnic results, unless something really unusual shows up. Like a large percentage from a continent that you weren't expecting at all.

                              The tests are pretty good at giving continental ethnicities, but not so hot on more exact geographic labels. Ancestry gives me 100% European with 52% Great Britain, 10% Ireland, and also amounts from Scandinavia and the Iberian Peninsula, with only a trace amount of 3% for West European. FTDNA gives me as 99% European with 62% West/Central Europe, 32% Scandinavia, 5% East Europe and a less than 2% Finnish (and on paper, I can trace back to a 9th G grandfather who was Finnish). So basically, the tests confirm I'm European, but geography is sort of up for grabs. I think of Ancestry as listing where a large portion of my ancestors were from, and FTDNA shows more where my ancestors' ancestors were from.

                              Comment

                              Working...
                              X