View Full Version : Shared segments-common ancestor or not?

09-09-2012, 01:45 PM
I need help! I am getting really confused by different comments from different people on other forums about how to decide whether autosomal cousin matches that share a segment with you and each other are really descended from a common ancestor. I'm hoping someone here will be able to give me a clear answer!
Here is my situation.
On a specific chromosome I have cousins A, B and C, all sharing a segment of roughly the same size. Cousin A and cousin B are parent and child. Cousin C is another individual who is also a 5th cousin to cousins A and B in this same place.
Cousin A and B also share this segment (not surprising as they are parent and child).
All 3 cousins got together and compared family trees. They are all part of a small community of people and they have found surnames and a town in common. I cannot compare as I don't know my ancestry.
I have been told these matches do not definitely prove we all share a common ancestor. Does anybody here know why this might be?
It is driving me nuts!
I also share a segment between Cousin C and her Granddaughter over at Gedmatch on a different chromosome. And with Cousin C, her son and his daughter on yet another chromosome.
Any help and advice would be most welcome as I want to get a proper understanding of how shared common ancestry actually works!

09-09-2012, 03:11 PM
what is confusing, is the possibility of facing a true IBD segment, which is unbroken either on your paternal or maternal side and not mixed together from both strands (IBS). But yours is looking good.
I made some illustrations about that:

09-09-2012, 08:35 PM
Thanks Zaender,
I like your visual representations! I think based on what you have said, they do triangulate because they do all share in the same place with each other, however can anyone tell me if I've got it right:

So me and cousin A and cousin B do share a common ancestor because A has passed on that segment to their son B (let's say via her own mother) and I match them both and vice versa?

However A, B and C could also share in the same location via cousin A's paternal line, meaning they all share a common ancestor but not the same one as me and A and B and I happen to share in the same location with cousin C but on a different branch?!

If anyone out there thinks I have got it wrong please let me know as I want to get this concept cracked once and for all...

09-09-2012, 09:01 PM
I can further add to this because I know in my case, that my cousin B is the son of cousin A so his segment is from his maternal side so no possible problems via his paternal line there. Cousin C is from the same community of people as cousin A.

I'm not an expert either and all visualisations are my private approach to a deeper understanding.
The main problem we encounter is called PHASING, the clear separation of mat and pat strands. Chip technology doesn't do it and all the other attempts like gedmatch did not yet convince me, because they use tricks.
But if genealogical comparisons give confirmation, the chance rises, that you face an IBD segment.

09-09-2012, 09:45 PM
Thanks. This is the only scenario I have to date, where so many people share a segment and have surnames/ a town in common. How on earth do you decide if a single match is IBD or IBS when you know so little of your ancestry?! Or even if you do know it and can't find a connection...

09-21-2012, 06:55 AM
Since A and B -- the parent and child -- both share the segment with C, it seems likely to me that this is a single shared segment -- IBD (identical by descent) and not IBS (identical by state). And you say that you share with cousin C on multiple chromosomes, and that you also share with descendants of C. So it looks to me as if C is related to A and B and to you, but that doesn't necessarily mean all of you have (recent) common ancestors.

When I was a child, I had a bit of a crush on a girl named Jane. We lived in different states and I only saw her when my family visited my mother's brother's family. His children were cousins both to me and to Jane and her sister, but on opposite sides. That is, they were related to me through their father and my mother (brother and sister), and they were related to Jane through their mother and Jane's mother (sisters). As far as I know, Jane and I were not related at all. Or at any rate, not recently related (within the past couple of hundred years, at least).

EDIT: I may have misread your post. If you're saying that you do also share with A and B, then I'd say it's more likely than not that all of you share a common ancestor.

09-30-2012, 08:30 PM
Hi geebee,
Thanks for your post. Yes I do also share with A and B :)

10-02-2012, 06:28 AM
Nice explanation, beegee. Her connection is almost certainly IBD then.

10-04-2012, 03:23 PM
Thanks for everyone's help-I found different set of cousins today who all share with each other and me :biggrin1:

10-10-2012, 06:48 PM
I have a new problem involving shared segments!
Dare I say it, on the centromere of chr 10 I have the following matches (names not real):
Bob, British
Mary, 100% german
John. Dutch/unknown
Peter Half German/half British.

Now, I have checked Bob's AF as we shared genomes and he shares with Peter and John but not Mary.

Checking Mary's AF, she shares with [COLOR="#FF0000"]Peter but not John or Bob!

What has happened there?! Mary only shares with me and Peter, not the others. So does Mary share with Peter on one side of Peter's family and the others on his other side and I am related to both sides?!

That's a real doozy!!!