The Shared cM Project - one long segment vs multiple smaller segments

Question

Good morning, everyone!

I often use Blaine Bettinger's tool for predicting relationships based on cMs: The Shared cM Project 3.0 tool v4 (I use the one in my DNA Painter account).

Today, while looking at the results for inputting 54 cMs, I began to wonder if the fact that these particular 54 cMs were in one segment (rather than, say, 3 segments totalling 54 cMs) should make me start with closer relationships being more likely than the further options.  Or vice versa! Or neither, and the odds are just what they are!

Any thoughts?

Cheers

Shirlea

Answer 1

According to Blaine Bettinger, a quote from his book "The longer a shared segment is, and the more segments two people share in common, the greater the liklihood that the two individuals share a recent common ancestor".

He says, and this is the WikiTree Triangulation rule, that 7.7cMs are required for a Triangulation. FTDNA uses 5cMs to match.

So, my interpretation is that longer segments and then multiple segments means a closer match than the chart might show. Only a 7.7cM segment with a match would be a more distant relative. However, he does say that there are instances where first or second cousins have a very low match on cMs and they are known to be that relationship.

I don't know if this helps your quandary but I was intrigued by your question as I have been doing some triangulation and it isn't always as easy as it sounds!!

Comments

Thank you, Susan for the Star.

Answer 2

Some thoughts:

Some regions of the genome recombine more than others.  Segments in regions that are less prone to recombination will pass through more generations intact.  This is especially true if there are many males in the line of descent, since fewer recombinations occur on average in the creation of male gametes than female.

There are also "pile-up" regions, regions where large groups of people match each other.  More generally, there are segments that are "identical by population", meaning there is some population of individuals that have low diversity in a certain region of the genome, so distant cousins will match in that region not because of a shared ancestral pair that can be nailed down but because they both have ancestors who came from that population.

So I'd check to see if this segment is in a pile-up region.  A list can be found here:

https://isogg.org/wiki/Identical_by_descent

Otherwise, just be aware that segments that long can sometimes show up in very distant cousins.  My mother shares a 45 cM segment with someone with whom I am reasonably confident both that the MRCA couple were both born around 1720 and that endogamy is not at work.  I have numerous long segment matches in pile-up regions with people with pretty deep trees and with whom I can find no inkling of a connection.  

Comments

Thank you for this information, Barry. So much to learn!!!

Answer 3

A single relatively long segment may pass down over many generations without recombination; I have a segment shared with at least 40 other people on GEDmatch, the largest triangulation group I have for a single relatively large segment, most of whom both I and my mother match for 30-40cM. The known MRCAs for most of these people are a set of my 6th great-grandparents, born c. 1740 in Northumberland, England; my relationship to most of these people is 6C1R/7C, and my mother's relationship is 6C/6C1R.