The question came up about siblings sharing the same crossover points. The answer is yes – some of them will be the same. Let’s look at this generation by generation. [There is often good discussion in the Comments to these blog posts – we are all learning on this journey. As a result of a recent comment, I decided to do a blog post about this topic]
The set-up:
1. One genome – let’s use our Mother’s side – 23 Chromosomes
2. Assume the average of 34 crossovers per generation.
3. A crossover is the point where DNA changes from one grandparent to the other grandparent, when the mother recombines her two chromosomes into a new one to pass on to a child.
4. Crossover points are random.
Mother’s DNA already has crossovers created by many of her Ancestors. She will recombine the DNA from her two parents at 34 places over the 23 chromosomes, and pass these new chromosomes to a child. Note: this means usually 0, 1, 2 or 3 crossovers per chromosome (on average 1 per 100cM). Since these crossovers are randomly formed for each egg, it would be rare for any of her children to have the same crossover from her.
The 34 new crossovers created 34+23=57 segments. These 57 segments “cover” all 23 chromosomes, from beginning to end of each one. These 57 segments are from Mother’s parents – our grandparents. All the crossover points from recombination events in prior generations are fixed (static) in the two grandparent’s DNA.
Example: Mother’s paternal DNA on Chr03 – from 47Mbp to 123Mbp has a crossover point at 68Mbp. Each of Mother’s children who got her paternal DNA that included the point at 68Mbp would include that crossover point. Mother could pass a paternal Chr03 segment 47-83Mbp to one child and paternal Chr03 59-119Mbp to another child – both of these children would have the same crossover at 68Mbp.
Note: The 68Mbp crossover could have occurred at the great-grandparent generation, OR at some previous generation.
This is a good example of why Chromosome Mapping *by generation* is important. In general Segment Triangulation results in Triangulated Groups (TGs) from different generations of ancestors. The TGs are not all from 4xG grandparents, or any other specific generation. However, if you have the Common Ancestor (CA) for your TGs, you can easily build a Chromosome Map for different generations. In my case I have 372 TGs – I know the CA side and grandparent for almost all of them – they roughly “fit” into about 114 groups (representing my 4 grandparents on both sides) on my 45 chromosomes.
Bottom line: Siblings won’t (generally) get the same crossover points from their parents, but likely will share some crossover points from grandparents and more distant Ancestors.
[05F] Segment-ology: Sibling Crossovers by Jim Bartlett 3 Dec 2022
segment-ology 
Hello Jim
Perhaps you can help me.
There is a connection between my uncle (brother B) and my father (brother A) that makes me feel uncertain.
I’m not entirely sure whether they are full siblings or 3/4 siblings. What made me feel uncertain was the low proportion of FIR.
In order to be able to analyze the whole thing a little better, I had further tests carried out, especially on other cousins of mine.
I’ll allow myself to give you a brief overview:
The parents of the two brothers A+B+(C); i.e. my grandparents; are distantly related; in the “3C” area
Initial situation: I still have the GEDMatch kit number listed
Brother A: YB029263C1 (my father)
Brother B: JJ7798703 (my uncle)
-> Analysis of Brother A+B: HIR 2772cM/FIR 641cM
Brother C (my second uncle): already deceased, no DNA available
Child1 of Brother A: FN9399793 (that’s me)
Child1 of Brother B: YS3247832
Child2 of Brother B: FU5808053
Child1 of Brother C: KQ7203396
When analyzing me and my cousins, I came to the following results:
Child1 of Brother A: FN9399793 to Child1 of Brother B: YS3247832
-> 971cM
Child1 of Brother A: FN9399793 to child 1 of brother C: KQ7203396
-> 906.3cM
Child 1 of brother A: FN9399793 to child 2 of brother B: FU5808053
-> 658cM
Rationally speaking, I would classify the connection between brother A and brother B as full siblings, DNA share approx. 48%
However, I am unsure about the FIR share and the result between me and my cousin, with a total value of 658cM.
I would be happy to hear your opinion or who I could turn to with the request
Thank you for your time.
Kind regards
Dave M
LikeLike
Hello Jim
Perhaps you can help me.
There is a connection between my uncle (brother B) and my father (brother A) that makes me feel uncertain.
I’m not entirely sure whether they are full siblings or 3/4 siblings. What made me feel uncertain was the low proportion of FIR.
In order to be able to analyze the whole thing a little better, I had further tests carried out, especially on other cousins of mine.
I’ll allow myself to give you a brief overview:
The parents of the two brothers A+B+(C); i.e. my grandparents; are distantly related; in the “3C” area
Initial situation: I still have the GEDMatch kit number listed
Brother A: YB029263C1 (my father)
Brother B: JJ7798703 (my uncle)
-> Analysis of Brother A+B: HIR 2772cM/FIR 641cM
Brother C (my second uncle): already deceased, no DNA available
Child1 of Brother A: FN9399793 (that’s me)
Child1 of Brother B: YS3247832
Child2 of Brother B: FU5808053
Child1 of Brother C: KQ7203396
When analyzing me and my cousins, I came to the following results:
Child1 of Brother A: FN9399793 to Child1 of Brother B: YS3247832
-> 971cM
Child1 of Brother A: FN9399793 to child 1 of brother C: KQ7203396
-> 906.3cM
Child 1 of brother A: FN9399793 to child 2 of brother B: FU5808053
-> 658cM
Rationally speaking, I would classify the connection between brother A and brother B as full siblings, DNA share approx. 48%
However, I am unsure about the FIR share and the result between me and my cousin, with a total value of 658cM.
I would be happy to hear your opinion or who I could turn to with the request
Thank you for your time.
Dave
LikeLike
Dave, the 658cM is still within the range for 1C. Jim
LikeLike
OK, thank you. Then we can assume that all three (A, B, C) are really full brothers, correct?
LikeLike
Dave – 1Cs share the same grandparents; and two 1C to you could each be from different grandparents. But in this case since they strongly relate to each other, they would be 1C on the same set of grandparents. Jim
LikeLike
hello jim, ok. I think the uncle/aunt to niece/nephew connection also points more to full siblings than to 3/4 siblings, so the cousin with the low value would probably be an outlier at the lower end, am I interpreting this correctly?
LikeLike
David – yes… you should go to DNA Painter and plug in the amounts, then click in the boxes in the Shared cM table to see the distribution curve. We have a specific cM number for each instance, but there is a pretty wide range to most of the distribution curves. We shouldn’t be concerned with a value that isn’t right in the middle. At this point it’s more important to try to determine the actual genealogy… I know it’s sometimes hard. But we cannot sharpen the data marshmallow any more that a distribution curve. Jim
LikeLike
Hello Jim
You helped me a while ago.
In this context, I came across a theory that interests me purely from a factual perspective.
When analyzing the topic of 3/4 siblings, I came across, among other things, the possible connection “Your father had a child with your maternal half-sister.”
At first, I thought that with 3/4 siblings, one parent always had to be your paternal or maternal uncle/aunt (grandparents remain the same, the circle remains closed).
However, in the variant “Your father had a child with your maternal half-sister,” a new genetic input would be added via the half-sister (the half-sister’s father), which would mean that one of the three-quarters siblings would have a new grandparent.
I have the following question:
– Why is this relationship also considered a three-quarter sibling relationship?
– Would this also be detectable in the X-DNA for this variant?
Somehow, I lack understanding in this context.
I would be very grateful for feedback.
David
LikeLiked by 1 person
David, Tough questions. I’m working on an intense project right now and don’t want to take the time. But I did write a blogpost about calculating the cMs in endogamy. As with all things DNA, there is not a firm, precise, answer. But to get my head around this, I always stop and think – every human being had one mother and one father – and each of them had two parents. The best thing is to forget about all the entanglements, and just draw the Tree for the one person: 2 parents, 4 grandparents, etc Some names might be repeated (it happens all the time, just usually not so close to each other). Good luck. Cycle this back in the fall, if you still need help. Jim
LikeLike
Pingback: Sibling Crossovers - Fort Worth City Local News
I am trying to recreate my grandparents’ chromosomes using an uncle and 2 siblings from a) my father and b) my aunt. Since I have a messy endogamous situation, I have been trying to use the cousins to improve my understanding of the uncle’s two chromosomes. I am looking for 4 TGs at any given point on any given chromosome, so the more diverse the cross-over points between siblings (cousins pairs) the better. (I think?)
LikeLike
Kathryn – a great project – it should keep you busy for months, if not years. I have atDNA for my father, my brother, and my maternal uncle, and have been able to determine many of my grandparents’ segments. But I only got 1/2 of my parents’ DNA, my brother adds about 1/4 more; but the final 1/4 of each parent’s DNA is lost. My mother’s brother provides about 1/2 of the DNA from my maternal grandparents – some of that overlaps with what my mother passed to me – some of it dose not.
In your case, you have some kits to work with. As I’ve noted, the crossover points in your grandparents’ DNA from prior generations are fixed. The DNA they passed to each of their children will add roughly 2×34=68 new crossover points to the mix – they are randomly different for each child. Ditto when the children pass this DNA down to your generation. Some will be diverse, some will not.
I would first form Triangulated Groups for each person who has tested and determine the side for each one. Then using genealogy, determine the grandparent for each TG. This is easy in some cases, hard in others. Build the Chromosome Map at the grandparent level – this will involve combining some TGs into larger “super TGs” for each grandparent. I have 372 TGs, and roughly 2×34=64 total crossovers in 45 chromosomes for 109 total segments at the grandparent level – roughly 3.5 TGs need to be combined for each grandparent segment. Your DNA would account for about 1/4 of each grandparent’s DNA map. Overlaying the grandparent Maps from the other kits you have may extend this out to about 1/2 “coverage” of your grandparents, depending on what the aunt and uncle got.
On the one hand we’d like to know more; on the ohter hand, the crossovers you can find will provide a very valuable background “grid” as you look for Common Ancestors with all of your other Matches.
Please report back from time to time on your progress… Jim
LikeLike
Jim,
What type of research question would you expect cross-over point information to be most useful in?
Thanks very much for the super-clear description of recombination and cross-over points.
Mark.
LikeLike
Mark, A good question. Here are some ideas:
1. A Quality Control (QC) check. If you do a Chromosome Map and have 10 or 80 crossovers at the grandparent level, I’d take a hard look at it. In a few cases I’ve had an unknown TG on one side (I wasn’t sure of the grandparent. But if I chose one, it gave me a zig-zag and two more cross overs in a short span; if I chose the other, I’d link two with the same grandparent who was on each side of the TG and the big picture wwould be better, too… I focused on the latter and eventually found two MRCAs with that alignment. A tool that focuses your research is good.
2. With one or more siblings, it’s another kind of QC. Did the sibling TGs align when dealing with the same grandparent side? Can I impute a grandparent side in one sibling based on “known” TGs in a different sibling?
3. Crossovers determined in a parent’s DNA can certainly be imputed to their children if they passed that part of their DNA to them.
I’ve pointed out many times that our own DNA is all *true*, and it’s also fixed (there is only one “soluntion – 6 billion base pairs that don’t change). This also applies to our siblings and parents and, actually, to all of our genetic relatives. So if we figure out something at a particular point in our DNA (crossover, or unique SNP, etc.), it will (ususally) be the same in every relative who also inherited a bigger segment of DNA (with that point in it) from the same Ancestor. It’s kind of cool when you think about it… I could also go off on a tangent and think about where, in geography, each piece of our DNA was in time – it traveled to us from Ancestor to Ancestor (from conception to conception). It’s five-c’clock somewhere – signing off:>j Jim
LikeLike
On Chr 7, I am FIR with my brother between 95-154 Mbp. I determined that my brother and I have the entire paternal Chr 7 in common. So, if the likelihood of my dad recombining his chromosomes at the same place is highly improbable, then that seems to tell me that those segments on Chr 7 all came from either my grandmother or my grandfather. Ie, we got a whole Chr 7, not a recombination.
LikeLiked by 1 person
So, once you determine which grandparent it was, that will help finding more Common Ancestors. Jim
LikeLike