Effect of cousins marrying cousins

[Chester]

It seems a few of my ancestors married 3rd and 4th cousins, including my own GGrandparents.

How does this effect my DNA results, or is there no effect from it?

Does anyone know?

Thanks.

[hfp43]

It seems a few of my ancestors married 3rd and 4th cousins, including my own GGrandparents.

How does this effect my DNA results, or is there no effect from it?

Does anyone know?

Thanks.

There seems to be some dispute among the experts on this, but several have told me that a history of cousin marriage, especially when combined with very large family size, can "convince" the matching algorithm that the relationship is closer than it really is. This is certainly the case with descendants of my own French-Canadian ancestry.

[192971]

It seems a few of my ancestors married 3rd and 4th cousins, including my own GGrandparents.

How does this effect my DNA results, or is there no effect from it?

My father is 5th cousin of himself, via a ancestor couple who seem to be unexpectedly reliable in connecting me to FF matches on my paternal side. But of course it has been a random shuffle; I, or anybody descending from the ancestral couple in question, still could have ended up getting insignificant amount of DNA from them. It only have had lower probability to happen.

The genetics word "recombination" itself gives us a clue that there must be an effect of what is recombined. If the replaced part is totally different to replacing part, recombination is efficient, but in the other end, if the replaced and replacing parts are totally identical, recombination is like it never happened.

[T E Peterman]

I am descended from a few cases where there were first cousin marriages. I have found at AncestryDNA, that nearly half of my leaf hinted distant matches come through these cousin marriages. It amplifies the signal; makes the algorithms estimate a closer kinship & thus "see" a lot farther. A distant cousin that you connect to through that intermarriage will be a lot more likely to turn up as a match.

I have been describing 1st cousin marriages. The effect would diminish considerably at the 2nd cousin level & even moreso at the 3rd or 4th cousin level.

Timothy Peterman

[LadyAlaise]

There seems to be some dispute among the experts on this, but several have told me that a history of cousin marriage, especially when combined with very large family size, can "convince" the matching algorithm that the relationship is closer than it really is. This is certainly the case with descendants of my own French-Canadian ancestry.

I am no expert in genetics but to me it seems to make sense; I also have a lot of Acadian French ancestry with many many many overlappings of 3rd-5th cousins marrying on many generations in many branches of my tree. (we acadian genealogists like to joke that we have Family Fishnets not Family Trees perse), another reason I am looking forward to getting my ff results I have many Acadian cousins waiting to see how closely we match due to having multiple different generational family tree connections (ie: SoandSo is my 5th cousins through X line and my 7th cousin through Y line and my 4th cousin through Z line) or another example; my paternal Grandparents, going Up their Trees from them, were 3rd THROUGH 11th cousins respectively to each other..

[hansonrf]

A few words to look up...

The phenomenon is called 'endogamy'.

The impact on DNA and relationship calculation is referred to as 'Founder Effect' or 'Pedigree Collapse'.

It occurs in many 'tight-knit' communities, most notably where religious beliefs or customs or other constraints led to marriages between cousins to some degree.

[Dawn]

I have quite a few generations of 2nd cousins marrying.
I have a man who had shared cM's of approximately 70. When we dug into our tree, I found out he's my third cousin 2 times removed. I'm assuming we should be sharing way less DNA than we do.

[crossover]

There seems to be some dispute among the experts on this, but several have told me that a history of cousin marriage, especially when combined with very large family size, can "convince" the matching algorithm that the relationship is closer than it really is. This is certainly the case with descendants of my own French-Canadian ancestry.

well in the case of my grandpa(whose parents are related in multiple ways and happen to be second cousins)gedmatch incorrectly predicted them to be within the 3rd-4th cousin range

[K. L. Adams]

My wife's, father has a couple of double first cousins. Young folks in little southern towns did not stray far, especially if they only had a mule It will be interesting to see how family finder works for this particular line.

[Adam Nisbett]

In general based on what I understand of how DNA recombination works, cousin marriages end up giving you a larger dose of the common ancestry (since they would show up multiple times in your family tree this makes sense) but it's more likely to show up in multiple segments rather than one long segment. If there's a repeated history of cousin marriages you may end up with lots of shared DNA that's all broken into small segments.

[hansonrf]

In general based on what I understand of how DNA recombination works, cousin marriages end up giving you a larger dose of the common ancestry (since they would show up multiple times in your family tree this makes sense) but it's more likely to show up in multiple segments rather than one long segment. If there's a repeated history of cousin marriages you may end up with lots of shared DNA that's all broken into small segments.

And recombination can make 2 small segments one larger segment again...[not necessarily x + y, but larger than either alone]

[Adam Nisbett]

And recombination can make 2 small segments one larger segment again...[not necessarily x + y, but larger than either alone]

Yes, that's also possible, but statistically I would think it much more common for the segments to fragment than to merge. So for any segment that ends up more intact there will probably be multiple fragmented pieces. Since we are talking statistics though, for any particular case it can easily go against the norm.

[mamoahina]

Yes, that's also possible, but statistically I would think it much more common for the segments to fragment than to merge. So for any segment that ends up more intact there will probably be multiple fragmented pieces. Since we are talking statistics though, for any particular case it can easily go against the norm.

Correct. It looks more fragmented, not necessarily merging and appears more like a comb with missing teeth.

[sarahmj]

I think this issue is affecting my results to a degree as well.
My mother's ancestry is from Malta which is a very endogamous population. Her paternal grandparents are 1st cousins and her maternal great grandparents are first cousins. Then after phasing my Mum's DNA with my Grandma it seems she shares about 32cm with the side my Mum inherited from her Father (can't get his DNA since he died a long time ago). So not sure how my grandparents are related but it seems they are cousins somewhere not too far back.
I do have a few matches where there are many segments but the size of the segments are smallish....so probably not a close match.

[mamoahina]

A few words to look up...

The phenomenon is called 'endogamy'.

The impact on DNA and relationship calculation is referred to as 'Founder Effect' or 'Pedigree Collapse'.

It still baffles me why people use founder population/effect and population bottleneck interchangeably.

A lot of populations start from founder's population which I've been told even by a population geneticist that a founder's is a population bottleneck which I understand. But my own population have gone through, apparently, several population bottlenecks and the most recent one was in the 1790s. It declined since but started to go back up after 1900.

I've also seen this with Pitcairn people and how in 1850 they went through a severe bottle necking even though they started with a founder's population in 1790.