DNA relatives

[MarcuccioV]

I’m having some conflicts (this is nothing new, my entire genealogy research thus far has been one big conflict) with my DNA relative matches (on ALL the different matrices I’ve uploaded to).

I have quite a few DNA matches that based on their own ethnic makeups, are 100% Italian (a small percentage of the whole, but maybe a dozen or so all told).

Since my mother is Italian (with some other Mediterranean ethnicities thrown in at smaller percentages), I would expect that. Right..?

Except NONE of them can be tied to my mother’s branches conclusively. Not one.

Furthermore (and therein lies the quandary), almost EVERY one CAN be tied to my father’s side (by way of shared matches with people that I CAN conclusively tie to my father’s branches.

So how can this be if I can’t find any DOCUMENTED Italian on my father’s side (I have suspicions, but no paper trails – I also cannot rule out illegitimate births or adoptions even though I haven’t found any)..?

If, and I repeat IF, the Italian is related to my mother’s side AND an unrelated Italian ancestor on my DNA matches family NOT tied to my dad but related to me through another ANOTHER ancestor on my dad’s side, wouldn’t that be a bit of a stretch to assume they are related to me through two (totally unrelated) parental lines..?

And since my mother was the first generation born in the US, wouldn't that even INCREASE those odds dramatically..?

This has not occurred once, but SEVERAL times. Anywhere from the 3rd-8th cousin levels. Some of these matches (2 or 3 at least) STILL live in Italy.

Do I have to throw in the towel and assume they are just coincidental matches (although the odds seem astronomical) or is there another explanation..?

I won’t even throw into the mix the HUNDREDS of matches all told that show at least SOME Italian ancestry (again ALL related to matches on my dad’s side), since I COULD theoretically assume they match through whatever portion of their DNA is NOT Italian (one example I can think of is <97% Italian and >3% English).

It also ties in with my ethnicities estimates which give me <50% Italian (yes, I know, I can’t rely on that, I’ve been chastised to the Nth degree on that already, we don’t need to go there).

But it’s STILL a mystery...

[darkerhorse]

Do you have a family tree online, or at least the direct line in question?

[MarcuccioV]

Do you have a family tree online, or at least the direct line in question?

Yes, both. So let's break it down to ONE example to simplify things. Let's take match "X" (100% Italian according to their ethnicity report). Since I am at least 50% Italian you'd think that match would be to my mother's (Italian) side, right..? Wrong. Their "shared" matches (with me, at times 6 or more) seem to have at least some Italian ethnicity (which makes sense) and are connected by varying degrees of relationship. Except they are paternal matches.

I hope we are on the same page (so far)...

Now I can understand "X" being a match to me, since I have at least 50% Italian. And I understand the shared matches being matches to me, as I can tie common paternal ancestors or ancestral surnames/places to them.

But both..? That would mean "X" would have to match my 'shared' matches by their common Italian DNA, but also ME from my MOTHER'S side (not connected to any of our shared matches since they connect to my paternal side)..? And if "X" lives in Italy, that makes the connection nearly impossible.

The only way to explain that is if my shared matches are related to BOTH my maternal AND paternal sides. And I think ANYONE would have to agree that that's a huge longshot. Especially when it happens several times.

UNLESS I can still find the "mystery" Italian on my father's side, I (at least) can't explain it other than coincidence or human/computer error in the matrices. I'm not a big believer in coincidences, BTW...

And just for good measure and to show I'm not trying to "manipulate" the data, NONE of the surnames in these Italian matches are found in my mom's tree except ONE. And he has 2% Germanic DNA, and of his 2 shared matches, 1 has 2% UK (the rest Italian) and the other has 38 % UK (the rest Italian) with a common ancestor deep on my father's side as well as several matching ancestral paternal surnames (plus she shows as a 1st-2nd cousin but neither of us have heard of one another).

It just has my head spinning is all...

[darkerhorse]

So, is your family tree accessible to the public?

[MarcuccioV]

Yes, tree is public. Some of the matches are from different matrices. Those are all private. I don't mind sharing my info, but not the info of others. That is their decision to make.

[afecad]

Lot's of possibilities on the population of matches someone would see, this all applies to Autosomal for the most part

-Number of siblings on Mother or Father's side that had offspring

-% of those people who have done DNA testing and share the data publicly, believe it or not there is a high % of people who keep their data private, that is cannot be found in X-matching.

-DNA variability inherited from Parents, although you inherit 50/50, you don't receive exactly the same % as a sibling, uncle or aunt and etc.. That is, two siblings would not have identical DNA ancestry reports.

Except for parents which you share exactly 50% of your autosomal DNA (except males don’t share their father’s X chromosome),there is no concrete accurate amount of DNA that you share with any relative. It varies and sometimes rather dramatically.

-If you are seeing cM matches of more than 200 cM then they are often close family, I see those people on mine but I know who they are, not real surprises here, beyond that less than 200 cM, this is where it starts becoming distant and often surnames are not the same. It only helps to see connections in this realm, if the people have good/accurate family trees to see the connection. As noted, on average people do not nor do they even know or list family names or any info.

Also since you cannot test your Mother on 23andme or Ancestry, you cannot separate her DNA from your Paternal side so it just groups them all together and the only way for you to know is seeing surnames, solid family trees or talking to people who know some history.

Gedmatch shows my closest relative Gen 4.0, that's totally useless in the sense of distance. Now Ancestry, FTDNA and 23andme, show somewhat similar results on matches, but there are people who show up on one but not the other, because they did not do testing at all three companies nor is there a one database for matching beyond Gedmatch, which doesn't show me anything.

It's plausible your Mother's family falls in some of these possibilities, though I'd expect to see some matches in Italy, I have a couple, even though my family left over 120 years ago, I know some siblings did not come over and stayed in Italy, which makes sense. Perhaps for you, the family line ended in Italy or wasn't big to begin with?

[MarcuccioV]

Lot's of possibilities on the population of matches someone would see, this all applies to Autosomal for the most part

-Number of siblings on Mother or Father's side that had offspring

-% of those people who have done DNA testing and share the data publicly, believe it or not there is a high % of people who keep their data private, that is cannot be found in X-matching.

-DNA variability inherited from Parents, although you inherit 50/50, you don't receive exactly the same % as a sibling, uncle or aunt and etc.. That is, two siblings would not have identical DNA ancestry reports.

Except for parents which you share exactly 50% of your autosomal DNA (except males don’t share their father’s X chromosome),there is no concrete accurate amount of DNA that you share with any relative. It varies and sometimes rather dramatically.

-If you are seeing cM matches of more than 200 cM then they are often close family, I see those people on mine but I know who they are, not real surprises here, beyond that less than 200 cM, this is where it starts becoming distant and often surnames are not the same. It only helps to see connections in this realm, if the people have good/accurate family trees to see the connection. As noted, on average people do not nor do they even know or list family names or any info.

Also since you cannot test your Mother on 23andme or Ancestry, you cannot separate her DNA from your Paternal side so it just groups them all together and the only way for you to know is seeing surnames, solid family trees or talking to people who know some history.

Gedmatch shows my closest relative Gen 4.0, that's totally useless in the sense of distance. Now Ancestry, FTDNA and 23andme, show somewhat similar results on matches, but there are people who show up on one but not the other, because they did not do testing at all three companies nor is there a one database for matching beyond Gedmatch, which doesn't show me anything.

It's plausible your Mother's family falls in some of these possibilities, though I'd expect to see some matches in Italy, I have a couple, even though my family left over 120 years ago, I know some siblings did not come over and stayed in Italy, which makes sense. Perhaps for you, the family line ended in Italy or wasn't big to begin with?

ANY of these are possibles. I understand the lack of matches maternally, as I am my mother's only child. She had one brother that survived to adulthood, who had 5 children, so I have 5 first cousins on that branch, and numerous 1st cousins once removed. My grandfather only had one sibling here in the US (older sister), who had 2 sons and a daughter, who all had children (2nd cousins once removed, I think). None of these come up as matches, meaning none have tested.

All other siblings of BOTH my grandparents remained in Italy and had families there.

The match we (both) can't figure out is a 1st-2nd cousin match (838 cM). She is 62% Italian, but has NO Italian surnames on her family tree. She is not aware of being adopted. She has many surnames in her tree that align with my paternal side, and a 5GGf match (on my father's side, not direct paternal line but an offshoot of one of the paternal line wives). So she is a complete mystery if her shared matches come from my maternal line (see next paragraph).

She does share 2 matches with me that are 98% Italian (and those are the ONLY 2 matches we share). One has an ancestral surname from deeper in my mom's tree (Latini) and is 2% Germanic European ancestry. I could match him on either side, then, but he has no trees. The other shared match is NOT a shared Italian surname, but 2% England & NW Europe so again I could match either way. No trees for her, as well. No luck in surname searches either, as I have no locations or ages. No common ancestors.

My main question was how could someone with 100% Italian match both me AND shared matches that I have common ancestors with on my father's side..? Not including the 98% example above.

I'm just going to accept it as either coincidence (which I am loathe to do) or corrupted data...

[afecad]

The match we (both) can't figure out is a 1st-2nd cousin match (838 cM). She is 62% Italian, but has NO Italian surnames on her family tree. She is not aware of being adopted. She has many surnames in her tree that align with my paternal side, but no exact matches to anyone (her tree doesn't go back as far as mine). So she is a complete mystery.

She does share 2 matches with me. One has an ancestral surname from deeper in my mom's tree (Latini) and is 2% Germanic European ancestry. I could match him on either side, then, but he has no trees. The other shared match is NOT a shared Italian surname, but 2% England & NW Europe so again I could match either way. No trees for her, either. No luck in surname searches either, as I have no locations. No common ancestors.

My main question was how could someone with 100% Italian match both me AND shared matches that I have common ancestors with on my father's side..?

I'm just going to accept it as either coincidence (which I am loathe to do) or corrupted data...

838cM is stronger than any of my cousins, but I know that most are First Cousin Once Removed (top one is 650cM), none of my 1st cousins did testing or share their results that much is obvious. 23andme accurately notes the nearest tested cousin correctly.

Have you pondered the possibility she is a child born out of wedlock and she doesn't know? That value falls in the median range for a 1st cousin based on the estimate chart below, but it could be First Cousin Once Removed as well.

shared_cm_DNA_chart.jpg

Since you cannot separate your father/mother with DNA, you don't know for sure what side it's on, beyond family tree data.

What % of DNA is showing shared on 23andme or is it Ancestry or ?

[MarcuccioV]

838cM is stronger than any of my cousins, but I know that most are First Cousin Once Removed (top one is 650cM), none of my 1st cousins did testing or share their results that much is obvious. 23andme accurately notes the nearest tested cousin correctly.

Have you pondered the possibility she is a child born out of wedlock and she doesn't know? That value falls in the median range for a 1st cousin based on the estimate chart below, but it could be First Cousin Once Removed as well.

shared_cm_DNA_chart.jpg

Since you cannot separate your father/mother with DNA, you don't know for sure what side it's on, beyond family tree data.

What % of DNA is showing shared on 23andme or is it Ancestry or ?

Ancestry. And she doesn't know, but I suppose it's possible. I did more research and found new information on one of the matches that did not come up on previous searches.

So here is the Menage a trois:

The 12% DNA 838 cM across 38 segments match (lets call her "Miss A") has the same exact 5GGf as me (same name, date of birth, birthplace, etc) on my paternal side. His last name is Overstreet. His ethnicity is English on both his paternal and maternal lines as far back as I can go (several generations).

Miss A's ethnic makeup includes some UK ancestry, as well as a majority of Italian. As stated, no Italian surnames in her tree.

We share TWO other people as DNA matches (let's call them "Mr X and Mrs Y").

So Mr X shares a surname with someone in my MATERNAL grandMOTHER's line.

Mrs Y has an unrecognized (married) surname. Further research gave me her parent's names. Her mother's maiden name is the same as my MATERNAL great grandFATHER's mother's maiden name (and she was born in my grandparent's home town of Valmontone).

Mr X and Mrs Y are ALSO DNA matches to each other.

So 'splain, Lucy.

How can a DNA match with a direct link to a common ancestor on my PATERNAL side have DNA matches to TWO unrelated people (to each other) that I can connect to both of my MATERNAL grandparents..? And ALL THREE are matches to me..?

The odds of that HAVE to be incredible..! But, there it is... Vegas, here I come.

[MarcuccioV]

Mr X and Mrs Y are 27 cM and 31 cM respectively to me, (not to mention the connections described above). I don't know how they compare to Miss A, but they are her ONLY matches. None of them (besides me) are on any other matrix for comparison.

[MarcuccioV]

One other thing, based on your attached chart (from Ancestry) and birth comment. 12% DNA and 838 cM matches puts her squarely in 1st-cousin territory. This would mean she could ONLY be the offspring of one of my 5 uncles (4 paternal and 1 maternal). And since I have no children myself, it can't be a grandchild, nor siblings, which eliminates a niece or nephew (by blood).

Her tree shows no common locations except my home state of CA. That eliminates 3 of the uncles (all paternal), as they never lived here or in any other of her ancestral locations. So it leaves ONE paternal and ONE maternal uncle.

As far as my mother's twin brother is concerned, the chances of her being his are about .0000000000000001%

As far as my father's younger brother, it COULD be possible, but how then is she a DNA match to not one, but BOTH my Italian MATERNAL grandparents..?

She also DOES NOT match my 1st cousin on my paternal side (my father's older brother's son) which I have an even closer match to (but within 1st cousin range). If she were my paternal uncle's offspring, she should be a 1st cousin match to him, too, but she is not a match at all. Weird.

I call that a mathematical equation that just doesn't add up...

[MarcuccioV]

So in further defoliating my Italian grandparent's branches, I actually find "Mrs X's" maiden name on BOTH my grandparent's sides. as a 5GGm on my grandmother, and as both a GGm AND a 3GGm on my grandfather (2 SEPARATE branches).

Even so, NONE of it explains the relationships to both my maternal & paternal sides...

[afecad]

Well the shared cM and % DNA shared can narrow down possibilities, but again this is only an estimate.

I only have 1st or 2nd cousin matches on 23andme & Ancestry, nothing on FTDNA. All the matches are known people and there is not question on relationship.

Giving the cM & % DNA shared, it falls into a probably first cousin once removed, but it could be another link. It's only an estimate but this might point to a 1st cousin having a child out of wedlock, but it doesn't mean it is either.

Just to put things into perspective on this "estimate", at the end of the day that's what it is and I know this person is my first cousin once removed, yet the estimates on Ancestry say that's only 26% probable with the given result of cM.

This is from my first cousin once removed via Ancestry:

update xmatch.PNG

Match 2.PNG

Then here, same person via 23andme, as you will see their prediction is spot on accurate vs Ancestry:

23andmematch.PNG

cM and % DNA are not the same, variation between both exists. So if she did get tested via 23andme, you may see similar results.

Other questions I would try to find out:

-Her age and geographical location during time of birth?

-Can the above place her (as a possibility) in vicinity of anyone in your family?

-Are her parents surviving and have they been tested? Does she have siblings?

[MarcuccioV]

Well the shared cM and % DNA shared can narrow down possibilities, but again this is only an estimate.

I only have 1st or 2nd cousin matches on 23andme & Ancestry, nothing on FTDNA. All the matches are known people and there is not question on relationship.

Giving the cM & % DNA shared, it falls into a probably first cousin once removed, but it could be another link. It's only an estimate but this might point to a 1st cousin having a child out of wedlock, but it doesn't mean it is either.

Just to put things into perspective on this "estimate", at the end of the day that's what it is and I know this person is my first cousin once removed, yet the estimates on Ancestry say that's only 26% probable with the given result of cM.

This is from my first cousin once removed via Ancestry:


update xmatch.PNG


Match 2.PNG


Then here, same person via 23andme, as you will see their prediction is spot on accurate vs Ancestry:


23andmematch.PNG


cM and % DNA are not the same, variation between both exists. So if she did get tested via 23andme, you may see similar results.

Other questions I would try to find out:

-Her age and geographical location during time of birth?

-Can the above place her (as a possibility) in vicinity of anyone in your family?

-Are her parents surviving and have they been tested? Does she have siblings?

I have contacted her to ask these questions (prior to seeing your post). I know ALL my maternal 1st cousins, and NO illegitimate children as I grew up with all of them for the most part. This seems to be the strongest match.

The fact that she DOES NOT match my known paternal 1st cousin makes those dots harder to connect.

All I can do is wait for her response and see if that shines any new light on the information. She is very curious as to her Italian side (based on prior communications), so she should at least be agreeable to mutual discovery.

She has yet to infer whether she was adopted or a foundling. I'm trying to patiently allow further info to trickle in...

[MarcuccioV]

Addendum: I only have one other 1st cousin match (on MyHeritage) which is the younger brother of the Ancestry 1st cousin match and known to me (sons of my dad's older brother). The next closest (also on MyHeritage) is a 1st cousin once removed (the daughter of my dad's youngest brother's daughter). Also known and confirmed. I have met them all in person.

The 1st cousin once removed has a 6.9% DNA match @487 cM and 17 segments.

The 1st cousin (from MyH) has a 15.3% match with 1084 cM & 26 segments.

The 1st cousin on Ancestry (brother of above) is 17%, 1212 cM @ 33 segments.