DNA indicates we are 2nd cousins but we have no matches in our family trees. Where do we go from here?

Question

One of the sites that has my DNA results, Geni, notified me of a DNA match with a second cousin.  I contacted him and we compared family trees up through our great-great grandparents and found no matches.  One of my great-great grandparents had a second marriage, so I was thinking that was the key, but that, too, was a dead end.  How can we pursue this further?  Would the details of our DNA match provide a clue?  Thanks so much for any help you can give.

Michael Hartner, Gardiner, NY

845-532-7143

Answer 1

Its possible that there was a parental relationship in one or both of your families that neither of you are aware of.

Look for other close cousin matches that you each share together.

Answer 2

I wasn’t aware Geni did tests, but googling, it looks like the are a vendor for MyHeritage kits. Is that right? Is that who performed your DNA analysis?

The label 2nd cousin is an estimate based on your shared DNA. If you are on the low end of the 2nd cousin range, you could be 3rd cousins, or even a little more distant. If Geni told you your shared cM and you share it here, it would be easier to determine the relationships that are reasonably possible.

Comments

My Heritage do own Geni so yes, I would say that Geni would be a vendor for My Geritage kits.

---

No, my DNA test was performed by Ancestry.com.  I uploaded the test results to Geni and a few other genealogy sites.  I'm not sure where my 2nd cousin had his DNA analyzed, but I could find out if it would be helpful.

MH

---

So how is it that you know you are second cousins?... your trees don't meet at a known ancestor. I assumed this meant you were reported to be 2nd cousins by the site where you had DNA tested. Do you both have tests on the same site?

---

Regarding the question of 2nd or 3rd cousins, both of us know the names, DOBs, and residences of our parents, grandparents, great-grandparents, and great-great grandparents.  So we've compared all that information (in case it's a case of 3rd cousins) already and have not found common members of our family trees.

On the issue of DNA segment, I believe that either Geni or one of the other sites may have identified the shared segment of DNA.  I just looked at our respective profiles on Geni and found the following:  

For Me:  "at"

For My 2nd Cousin: DNA Markers:  "Y TM70  mt H40    "at"

---

This still doesn't explain why you think you are 2nd cousins. What leads you to that conclusion?

The information from Geni you just provided states only the types of tests you took, not any information about the test details or how you match each other. You both need to have autosomal tests (the "at" you saw at Geni) uploaded to the same site where you can compare them before someone could give a confident answer to your question.

---

Well I guess it's good, then, that both our tests indicate "at."  I was notified by Geni that they've found a 2nd cousin on the basis of DNA results.  That what's started this.  I'd tell you exactly what it says but I'm having trouble locating a copy of the notification.

---

Okay, yes that's definitely just an estimate. And it is possible that the connection could go a little beyond 3rd cousins, but not far, and it's not very likely. Since you've checked through your g-g-grandparents, then Dennis's comment seems relevant. A good way to identify an NPE (used to mean "not the parent expected", now means "non-parental event") is to look at residences of all of the people in your tree and find two people who were very close together at the same time. Otherwise, you can mine your DNA -- look at the people who match both of you on the same sections of DNA where you match each other. If you can find such people and they have enough information posted to build out their trees, you might find a common ancestor to yourself or to your match. That would be the golden ticket to prove which one of you has an NPE in the tree.

It is best if you can get precise data about how you and this cousin match -- which segments of which chromosomes, and how long they are. I don't know how the Geni/MyHeritage partnership works. It may be that your kits are in the MyHeritage system already. If not, you can download raw DNA data from all places I know that do tests, so probably there is a way to get your data from Geni. You could then upload to MyHeritage and/or GEDmatch and/or FTDNA. If both of you have kits on any one of those sites, you can then get precise data about how you two match each other. If you can give us the total cM you share, it could help eliminate the possibility that your match is beyond the third cousin level.

---

Thanks.  If I can find the common cM, I will share it here.

Answer 3

I've been a little skeptical of some DNA results. I recently did a 23 and Me genetic test. The online report says that I match as a first cousin to two people that I know is patently false. I'm not so sure about the accuracy of some of these tests. I believe that it might help, and has helped, in certain situations. But they don't seem to be perfect either.

Comments

First cousins through full siblings would average around 12.5% shared DNA although several other relationships (for example, uncle through half-brother or great-uncle) have the same amount.

You may not be first cousins with your matches but the overwhelming likelihood is a close family relationship even if it contradicts the official narrative.

---

Eric, before you dismiss the results you need to consider the possibility of an NPE (non-paternal event).

I read last year this from a DNA study (names changed to protect privacy):

Through the FTDNA database, we were made aware of a Mr. X whose genetic signature at the Y-37 level is identical to that of one of our Patriarchs. As Mr. X's surname is not the same as ours, this suggests and "extra paternity event."

As a fascinating study combining DNA analysis together with public records, one of our members has discovered that one of our ancestors served in the Texas Rangers at a camp very close to where Mr. X's great-grand parents lived at the time that Mr. X's grandfather would have been conceived suggesting a liaison between our ancestor and Mr. X's great-grandmother.

---

DNA doesn't lie, people lie.

---

Hi SJ,

Just thought I would point out that the Y DNA result could quite easily show a larger number of variants if they tested beyond the 37 STR marker level.

I have a match at 67 with one variance which then increases to 9 variances at the 111 marker level. So that match goes from possibly being within the last 200 years to being about 500 or more years ago. Which is around the time that surnames started to be hereditary.

---

When you do a deeper search (111 vs 37), you have a much wider allowance for mutations.  In this case it was a 0 mutation match with no relationship at all to the paternal relatives in the previous surname; not likely to be in the 200 to 500 year time frame..

---

How many cM?

And do you know how they are really supposed to be related (which is to say, do you know that they're related in a way that is NOT 1C), or are you saying that you didn't know they were related AT ALL?

There isn't a whole lot of "play" at the 1C level, but there IS more than one possible relationship. Assuming nothing rare and unusual is going on, of course, or there's a certain amount of endogamy.

---

Basically, a half-niece or half-nephew would have the same level of cM as a 1C. Exactly what you'd see if a sibling had a child put up for adoption. If that sibling were male, even he might not know about this person.

Same deal applies if one of your grandparents had the same situation, producing a biological half-aunt, or half-uncle.

---

Hi All,

I haven't reached out to this person, who is supposedly a 1st cousin. However, all my first cousins are well documented. From what I can tell, this person is at least 10-20 years older than me. Both my parents were the oldest of their siblings. My sister is 9 years older than me, and anyone older than that would have been impossible (due to my parents ages). I don't know the cM overlap, as that data is not currently available. 23 and Me are reporting "11.2% DNA shared, 18 segments". Admittedly, I'm only just now learning about DNA. But this person can't be a 1st cousin to me. Perhaps some other relationship, but even then, this would be someone completely unknown, to a pretty well documented tree.

---

Eric, you need to use DNA to confirm your ancestors. You can see how here: https://www.wikitree.com/wiki/Help:DNA_Confirmation

This person is very closely related to you, whether a first cousin or possibly an unknown aunt or uncle, etc. If you didn't want to discover any kind of relationships like this, then don't do DNA tests. That's all I can say. There is an NPE or adoption in one or the other of your trees. 

---

Hi Jessica, I don't dispute the genetic match, and perhaps there is an NPE I'm not aware of. I'm just disputing the "1C" label as that is highly improbable. If there is some other possible relationship that could account for the degree of the match then that would make more sense.

I suppose it's time to do more digging.

Thanks for everyone's help and comments!

---

Eric, have a look at this page:

https://customercare.23andme.com/hc/en-us/articles/212170668-Average-percent-DNA-shared-between-relatives

Based on ~ 11% this person first fits into the most likely category of:

1st cousin
Great-grandparent
Great-grandchild
Great-uncle/aunt
Great-nephew niece
(4-23% with average DNA shared of 12.5%)

This percentage could also fall into these categories listed by the site:

1st cousin once removed 
Half first cousin
(2-11% with average at 6.25%)

---

Thanks, SJ. That makes a lot more sense.

I've sent a message to this person, and another one that shows up as a close relative as well. Both are unknown to me. We'll see what happens.

---

Best luck on first contact!

---

Eric, my known first cousins who've taken DNA tests matched at the following amounts with me: one shares 11.2%, the second shares 10.5%, and the third 11.3% of their DNA with me.

My first cousin once removed (my father's first cousin) shares 8.42% with me.

Take a look at your matches. Do you match on the Y chromosome? Do you have partial matches on the X? Do you share mtDNA?

As the relative a lot of people would rather not find, not because of anything I ever did but because of the circumstances of my birth, I find the attitude some people have towards unexpected relatives at once frustrating, amusing, and offensive. Whether you realize it or not, the attitude comes across as "No one in my family would ever give up a child" or "No one in my family would ever cheat on a spouse". There is not a person in this world who can completely know what another person can or would do. Most of us never really know ourselves, much less our parents, grandparents, aunts and uncles and every foible, decision, and temptation they ever had. 

---

Hi Jessica,

Very interesting. The two people that I match as 1C are sisters, I found out. And I match at exactly the same percentages you listed (11.2% and 10.5%).

I'm less concerned if there was a NPE. People are human and have flaws. (I already know of some skeletons in the family closet anyway.) What I'm really perplexed on is figuring out how this is even possible. My father is the oldest of his siblings. My mother is the oldest of her siblings (one other sister). My oldest sibling is my sister who is (was; she's passed) 9 years older than I am. She was born when my mother was 16 and my father was 18. One of the persons that I matched on 23 and Me is 13 years older than me. All of my (known) first cousins are younger than my sister. The math just doesn't add up for me. If I go to my grandparents ages, it also doesn't look very probable. My paternal grandparents were 25 and 20 years old when they had my dad. My maternal grandparents were 21 and 19 when they had my mother. Perhaps its possible that there is some unknown older sibling of one of my parents. It's more probable if it was on my father's side, just based on ages.

Anyway, I'll keep looking into this more.

Answer 4

Have you uploaded your DNA to Gedmatch?

Since you did your tests with different companies, the only REAL way to compare them is for both of you to upload your RAW DNA data to GEDmatch!!

Comments

Thanks, Robynne.  I will arrange that.

---

Can't hurt!

Answer 5

Nobody has mentioned ftDNA Big-Y. Since you are both males, it will be a great help if you can afford it, even more so if other more distant cousins have tested. I had an exact match at y-37, but at BigY level he proved quite distant. Not very accurately resolved as we have insufficient testers. I have found several likely people, but regrettably none are interested in testing, so good luck if you go ahead.

At atDNA level, MyHeritage are now showing triangulated links which look interesting but I can't understand them yet.

Comments

That would only work if their connection happens to be along their paternal lines, right?.

---

Thanks Frank. Good point regarding the male line only which I assumed as Y is mentioned elsewhere in the string tho not by Michael. If they are, then BigY would help as the guideline is one mutation every three generations. My 2c and I show this perfectly, but we need a sibling or 1c to test to determine our terminal SNPs. We vary by one at STR DYS 710 35>37.

Answer 6

You are likely looking for an adoption, illegitimate child, or other NPE in one of your histories.

As others have said, upload both sets of DNA (for free) to GEDmatch, MyHeritage, and FamilyTreeDNA. The first two will give you very good DNA match details, and all three may give you some common matches to cross-check, to narrow down which side(s) of the family you are related on.

If either of you know any close family who have had DNA testing, ask them to also upload to GEDmatch and do One-to-One comparisons with them. If there is no match where there should be one, that will narrow down where the break is.

Comments

The cM number of this match would be a good start though.

---

Frank,  I think I saw the cM numbers for both me the alleged second cousin on one of the countless pages I reviewed.  I didn't realize it was key and therefore didn't make note of it.  What can the cM number tell us?

Thanks.

---

Thanks, Joseph.  Your comment is helpful.  We'll work on this.

Answer 7

Hi Michael!  Thanks for joining our WikiTree!

Just to remind you, your autosomal DNA test can have matches on any of these surnames, including the question marks:  https://www.wikitree.com/treewidget/Hartner-8/10

Comments

Thanks, Kitty.  We've already reviewed those surnames and quite a few more beyond that.  No luck.

Answer 8

My only suggestion here is to look at the trees again, but this time, only pay attention to the places.    See if there is a town or county that the two families have in common, especially if both are there at the same time..

In my family there is a rather close dna match on ancestry between my family and that of another.  While I could find no matches between the trees at all, I did find that both our families lived in the same small town in Iowa at the exactly the same time.    That led me to the conclusion that somebody was fooling around.

I have another example of that sort of thing in my family as well.  Two families, where there is a dna connection, but no connection in the trees, but they are both in the same place at the same time.