Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
At what level of cM should a DNA match be considered unreliable for the purpose of comparing shared matches?
For context, I have a brick wall with my great grandfather. Portuguese naming conventions of the 1870's do not follow the "inherit the father's last name" style. The name he went by in the US may have nothing to do with one or both of the parent's surnames. The earliest record I have of him is the 1910 census and I cannot find any immigration documents or any family records from the Azores. From this standpoint it is nearly impossible to connect the dots on name alone.
So I've decided to switch to DNA, build out the trees of my more distant DNA matches and see where that leads. I've looked into the Leeds method and it makes sense from a grouping standpoint. All of the cM ranges in the Leeds method (90-400 cM) lead me to my grandparents. I need to apply that methodology to more distant relationships to find out which match aligns with which great grandparent.
I know this is part art, part science but knowing how DNA is distributed, I imagine at some point the relationship is so distant that two individuals could be related to me, but not match to each other because the three of us did not inherit certain DNA. So my question is, at what level should matches be disregarded for these purposes?
I have my great grandmother's lineage back to the 1600's so I am able to trace many of my shared DNA matches through those lines. My closest DNA match who does not match to my great grandmother's line is 36 cM. Is this a strong enough match to consider that this person should be classified as my great grandfather's line?
Hope my question makes sense, but if there is any additional info I can provide please let me know!
At what level of cM should a DNA match be considered unreliable for the purpose of comparing shared matches?
It's a little more subjective than having a clear cut off point. ISOGG tells us that any segment (not total) below 15 cM has some chance of not being identical by descent: https://isogg.org/wiki/Identical_by_...sitive_matches
But the likelihood drops as the segment gets smaller. The first significant drop (more than a few percentages) happens between 9 and 8 cM when it drops from 80% to 62%, but it's still above 50%, so there's still a greater chance of being identical by descent than not. At 7 cM it drops below 50% - down to 42%. This is probably why Ancestry's threshold is 8 cM. But you could still argue that 42% is not insignificant, and so there are other companies that have lower thresholds.
But if you want to be 100% sure a segment is identical by descent, the threshold is 15 cM.
Of course this is not accounting for high levels of endogamy.
Quote:
I know this is part art, part science but knowing how DNA is distributed, I imagine at some point the relationship is so distant that two individuals could be related to me, but not match to each other because the three of us did not inherit certain DNA. So my question is, at what level should matches be disregarded for these purposes?
Firstly keep in mind that if you're using Ancestry.com, the Shared Matches list only goes down to 20 cM. So two matches could actually match each other but don't appear as Shared Matches.
But even when two matches don't match each other, I don't think it necessarily means they should be disregarded altogether. If they both match the same other people to form a larger group where everyone matches most of the others in the group, it's usually safe to say they are descended from the same ancestor/branch.
Quote:
I have my great grandmother's lineage back to the 1600's so I am able to trace many of my shared DNA matches through those lines. My closest DNA match who does not match to my great grandmother's line is 36 cM. Is this a strong enough match to consider that this person should be classified as my great grandfather's line?
A total of 36 cM probably has at least one segment that is long enough to likely be identical by descent. But I'm not sure why that means they must be from a particular great grandparent - you have 8 great grandparents.
In general, I would not assume that not matching people from a certain branch means they are definitely not from that branch. Especially at Ancestry where there are limitations on the Shared Matches list. I only confirm a match is from a certain branch if the matches they have in common with me can be identified. I do not group people by process of elimination since that could be misleading.
Thanks for the response. I agree, endogamy is something that is always in the back of my mind. My ancestors come from an island and if you go back a few generations you can make connections amongst families. Not to derail my own post, but along the lines of endogamy is this something that can be ascertained through DNA or is it something that is only apparent once you build out a tree and make the connections yourself? That is to say, if I have a relative who, on paper, I discover is a 2c match, are there any factors in the DNA match that would indicate that I'm both a 2c match and a 4c match (or something along those lines)? Meaning we share a great grandparent from one line and 3rd great grandparent from another line, or is this something not readily apparent given the basic DNA testing by Ancestry?
Back to my original question, my goal here was not so much to eliminate certain people but rather focus on the most likely relatives that will lead me to my great grand father's line. These matches that I have on Ancestry generally have a tree built out to their grandparents (or perhaps on a line that is not the line I'm researching). So from that standpoint it is a huge time commitment to build out someone's tree to try to find a connection. My thinking is that if I can focus on those DNA matches who do not match to those on my great grandmother's line then I'm at least focusing on a piece of information that I don't have yet.
Alternatively, if you have suggestions that are smarter, I'm all ears. As it pertains to my great grandfather, the paper trail has been exhausted with the names and dates I have, so I'm trying to figure out how to use DNA to my advantage. I don't know enough about DNA to say with certainty "this is what I should be researching" so that's why I'm reaching out here.
The Azores have a highly endogamous population. Your matches from there likely share a much higher amount of DNA with you than you would expect for the relationship. Most of them probably do share multiple branches of their family trees with you, but there is no way to separate that from DNA alone. People related to your great grandmother may also be related to your great grandfather.
When you say "My closest DNA match who does not match to my great grandmother's line is 36 cM", do you mean the match's tree doesn't connect to your great-grandmother's tree? Or they are not a match "in common" with the ones you've identified from your great grandmother's ancestry?
And when you refer to your great grandmother's lineage, is it just one line, or her entire tree going back to 1600 (which would be unusual)?
The Azores have a highly endogamous population. Your matches from there likely share a much higher amount of DNA with you than you would expect for the relationship. Most of them probably do share multiple branches of their family trees with you, but there is no way to separate that from DNA alone. People related to your great grandmother may also be related to your great grandfather.
Correct and that is why I had the idea to seek out people who I DNA match with where we do not have shared DNA matches that flow through my great-grandmother. Thus (in my mind at least) isolating those who descend from solely my great grand father's line. But perhaps that's flawed logic.
Quote:
Originally Posted by aries63
When you say "My closest DNA match who does not match to my great grandmother's line is 36 cM", do you mean the match's tree doesn't connect to your great-grandmother's tree? Or they are not a match "in common" with the ones you've identified from your great grandmother's ancestry?
What I mean is that me and that match do not have shared matches who connect through my great grandmother. The shared matches here are unique to me and that person. But I guess my question is whether this means this individual descends from my great grandfather's line or just that our match is so remote, and DNA distribution being what it is, means that we just didn't get the same DNA as the others.
Quote:
Originally Posted by aries63
And when you refer to your great grandmother's lineage, is it just one line, or her entire tree going back to 1600 (which would be unusual)?
It is her direct line. Parents, grand parents, great-grandparents and so on. If I happened to find other relatives during the research they were added, however it is mostly just her direct line. The Azorean records are not indexed so it is a very manual and time consuming process to build out an entire tree. It's one of those things I'll get to when eventually, however right now my time is focused on at least getting some crumbs of information to follow to find my great-grandfather.
To put some numbers to it, these are the shared matches I have with this individual.
The one outlier is the first person who does in fact connect to my great grand-mother's line however that could be a one off due to endogamy in her direct line. This person is a super close match but won't give me her parents' names because she's new to Ancestry and doesn't feel comfortable.
Sorry for the multiple posts. One other thing to add.
This is my methodology I've used so please feel free to poke holes in my analysis. I have hidden names on the below snippet.
The top row represents DNA Matches/Relatives who I have confirmed are from my great grandmother's line. The column on the left are DNA matches and whether they share a DNA match with the person in my great grandmother's line. This was the basis for my posts above on whether someone flowed through my great grandmother or not.
An interesting side note to this analysis is that all of the people who have "No" across the board are located in Hawaii. The rest of my Azorean family came to Massachusetts or Ontario Canada. The person in the list who is all "No" with a 36 cM match is the tree I'm currently working on and the shared list I provided above. She is the closest match who meets the criteria and had a tree for me to base my research off of.
Sorry, but trying to help you with this is like back-seat driving, since I'm not sure I'm interpreting some of your statements correctly, and don't know who you're referring to when you say things like "this individual" or "this person" which is clear to you but can be ambiguous to someone else.
How many cM do you share with your closest match? Is it the 183 cM match or is that just the top match you have in common?
Just some general observations:
Not knowing the other branches of your great-grandmother's tree is a big blind spot. Matches who do not match your confirmed relatives through her researched line may be related to her other lines you haven't researched. It may not indicate that those are relatives of your great grandfather.
Also with endogamy, I think the matches under 60 cM are probably useless. These can be very distantly related. I don't have experience specifically with Azores but for other island groups I've encountered in DNA, they are too distant to be useful. Seeing a lot of matches in the 40-60 cM range is probably only the effect of endogamy, and not an identifiable shared ancestor.
Your matrix of "yes" and "no" matches to the confirmed relatives I think is only showing you the randomness of DNA matching, not giving you a picture of who may be related to your great grandfather. For lower-level matches it's a crap shoot.
The Hawaiian matches probably only indicate that at least one of your distant relatives from an earlier generation settled in Hawaii. It doesn't indicate which ancestor they are related to.
Do you know if your two great grandparents had siblings, and if they had descendants? That could be a way to help sort matches. I have an analogous situation with my two German-born great grandparents. I have some 3rd cousins in DNA tests who descend from siblings of my great grandfather, and some 3rd cousins who descend from my great grandmother's siblings. I can use these to find matches in common. Initially I didn't know who the parents or siblings of my great grandfather were, but once I found matches to 3rd cousins from that line (also descendants of immigrants to the US), I was able to find clues and fit the pieces together. So, either you might find some of these matches through DNA, or you can track them down to see if they would be willing to test.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
