MyHeritage DNA Matches

Question

Is it possible that a DNA match on MyHeritage might not show up if I search for it, because the person who took the test marked it as private? One of my matches told me that her mother also took the test on MyHeritage, but I do not see her mother when I search under the mother's maiden or married name. After researching my match's father's ancestry, it seems highly likely that I am not a match to him.

Answer 1

I don’t know of private kits. Did you also check the list of common matches for you and this match? If the name was misspelled or the search feature is wonky, the mother should still show up as a common match if it’s a real IBD segment that came to her from her mother. How strong is the match?

Comments

Thanks, Barry, for the quick response! Yes, I did check the list of shared matches between the daughter and me, and the mother is not on it. But the mother does have a family tree on MyHeritage and it is under the mother's married name. So, I am assuming that if the mother had DNA test results, it would be under that name. (But I guess there's no way to know for sure.) The match is 43.0 cM and the longest segment is 38.0 cM.  The mother has a grandmother with maiden name of Cox. The father has some relatives with last name of Cox, but no direct Cox ancestor that I have found yet (I have found all of his great-grandparents and 10 of his great-great-grandparents so far.)

---

So, a key question is whether MyHeritage DNA will tell you if a person has taken a DNA test. On Ancestry, it will say "Person X is either not a DNA match or has not taken a DNA test." Does MyHeritage provide this or similar info?

---

I don't know of a feature at MH like the one you mention at Ancestry.

43cM at most sites would be far too large to be a false match. MyHeritage uses imputation, which I have seen cause false matches up to 20 cM or so.

However, there are a few regions where MyHeritage has known problematic matching. I am think of the beginning of chromosome 15, which is causing my uncle's kit to be overloaded with matches that don't triangulate, so must be false, with a few of those matches over 40cM. So out of curiosity, can you tell me what region of which chromosome the matching segment with the daughter is on? Maybe look through other high-ish matches and see if there are others who match on the same segment but do not triangulate. If so, then it may be another one of these issues with the algorithm at MH.

Otherwise, here is a cautionary tale about dismissing the matches on the paternal side so quickly.

I also have a mysterious group that parts or all of a 45cM segment. This seems to be what people call a "sticky segment", although I don't necessary believe such exist separate from the occasional segment that breaks through the constraints of typical probabilities, by random chance. One group of 8 matches on this segment trace back to a couple born in New Jersey in the late 1700's, two groups of 3 matches trace back to the early 1800s in Tennessee, and another group of 3 matches trace back to the early 1800s in Pennsylvania. I can't yet find a common ancestor between any pair of this groups, but I built the groups themselves by finding common ancestors between the testees in the groups.

This is far beyond what I am used to. Mostly common ancestors present themselves, or they don't, but I've never in another case had so many individually triangulating groups that seem to spring from so far in the past. I sometimes wonder if it is common to have, say, one group of matches like this.

 The segments triangulate, so I am convinced there was an ancestor back maybe even in the 1600s who sent the segment down to all of us. But if your paternal ancestors were on separate continents with no hope of separate but convergent migration, then that would add more certainty to the match being on the maternal side.

---

The match is on chromosome 5. I share this same segment with a third cousin (call her R.E.) with whom I share a pair of great-great-grandparents on my Cox line. R.E.'s granddaughter also matches me on this segment. So, I am investigating R.E.'s and my shared matches. Other than R.E. and her granddaughter, I have 7 matches on this same segment. So, I am researching them in hopes of making a deeper connection on my Cox line.

---

I don't think it is a false match. The problem with the mother's line is that the Coxes there don't share common locations with my Coxes. Hers are Kansas, Nebraska, and Indiana and mine are Ohio, Texas, and North Carolina.  (In that order going backward.) But the Coxes in the father's family (although they are not direct ancestors) are from South Carolina and North Carolina.

---

Kansas and Nebraska were settled from the mid-nineteenth century. Indiana mostly from the early nineteenth century. But South Carolina and North Carolina were settled 100 years earlier. The match could be on the father's side in the 1700s in the Carolinas. Or maybe elsewhere, say Virginia, with people later migrating to the Carolinas.  So I wouldn't fixate on the places.

I have a 35cM segment in common with a match where I am confident our most recent common ancestors were both born aroudn 1720. Yours could definitely be coming from the mid 1700s, or even earlier.

Answer 2

I'm not sure about privacy settings, however, I see a lot of people using initials only or nicknames, particularly when the account is managed by someone else.

Comments

It's a good point. That may provide additional privacy. But I don't think that is the case here, because the mother and daughter both submitted a very detailed family tree.

---

On the subject of sticky segments: On MyHeritage, I am looking at a case where a father, a son, and his grandson all match me and a third cousin on the same 20.7 cM segment on Chromosome 1. Is this an example of a "sticky segment" that you mentioned above, or something that is common? -- Thanks, Robert