cousinship and longest block

[stennor]

I'm writing to share what I'm observing regarding the correlation between the longest block of shared chromosomes and documented cousin relationships, and to ask what others are observing.

We're led to believe that the longer the longest block, the closer the relationship. However, in general, the correlation appears to be low, so it's problematic to rely on the longest block to predict cousinship. Here's what I have on my list, showing the longest block and the documented relationship:

8.55 cM tenth cousin
8.61 cM sixth cousin
9.09 cM seventh cousin
9.42 cM tenth cousin
22.68 cM ninth cousin and tenth cousin
28.41 cM multiple cousin relationships: sixth, ninth, ninth, tenth and tenth
29.41 cM tenth cousin

Now, there could be a number of reasons for the lack of correlation. Regarding my tenth cousin who shares a block 29.41 cM long, there could be a closer connection of which we're unaware. Regarding my sixth cousin with 8.61 cM, it's possible that we simply don't share much DNA.

I have yet to find any connections as close as those that Family Finder indicates (e.g., a range of 2nd-4th cousin with a suggested relationship of 3rd cousin).

In part, what I'm observing reminds me that recombination is indeed random; that is, the DNA doesn't pass down evenly and predictably through the various branches of the family. As many on this forum already realize and as FTDNA explains, it's possible for a pair of first cousins to see contradictory results regarding someone to whom they've documented a common relationship: that is, I show a match to a documented sixth cousin, but my first cousin, who is related to that person in exactly the same way, doesn't show a match at all. I share 50% of my ancestors with my first cousin but less than 50% of my DNA.

All of this is leading me to be wary of assuming a certain degree of relationship on the sole basis of shared chromosomes. I found a more expert explanation of some of these questions in a post by a geneticist from the University of Maryland: http://ongenetics.blogspot.com/2011/...e-segment.html

Any comments?

[blejerh]

This is a much needed reminder, thank you. My sister's results came back recently and they were compelling proof of recombination.

I share 31 cM with a 6th cousin. My 3rd cousin and his daughter also have longish overlapping segments shared with that individual in the same place. Yet my sister shows barely any shared DNA with him there.

Contrast my 5th cousin who I have only 10 cM shared with. And people who I have not been able to document any shared ancestry with who share between 25 and 30 cM with me.

There are also specific segments on specific chromosomes which a large number of us Ashkenazi Jews share (e.g., Chromosome 12), so many of us that I have to think the share ancestry is much further back than indicated.

[travers]

Those are all small segments. That's why making predictions of relationship is so difficult. I've never really heard that it's about the longest block. 2nd cousins and much closer relatives yes. If you've ever tested close family than that is where the large blocks come into play. 3rd cousins and further not so much, you will generally find smaller segments such as what you are finding.