Segment data of DNA matches - useful or not?

[PhilGee]

Shortly after getting my Ancestry DNA results, I uploaded the data to to GEDmatch, MyHeritage and FTDNA (just because I could). These all identify the chromosome and the segment genomic position for matches.

So, does knowing this additional data add useful information to the pot or is it just window dressing? I'm thinking mainly of the dreaded "brick wall" but not limiting it to that.

Phil

 

[jorghes]

I'm not sure that for the lay researcher not up with the science behind DNA, that segments and the like are really important.

Does it matter if you don't know where that person fits anyway?

 

[peter]

If you know how and the cousin who shares that segment are related then you can attribute that segment of DNA yo a specific ancestor (or, more usually, couple). But whether that information is of any practical value is a different matter!

Where you don't know how you're related you can identify whether the segment is in a pile-up region. If it is, this may well explain why you've been unable to find the connection.

 

[Pauline]

While the 'shared match' feature at Ancestry provides some of the useful information chromosome browser tools might provide, its big limitation is that it only includes 'close cousin' matches. It would perhaps be more useful to be able to extend the scope of shared matches to a user-defined cM limit, but since you can't other methods of determining 'shared match' groups can sometimes be useful.

 

[PhilGee]

Where you don't know how you're related you can identify whether the segment is in a pile-up region. If it is, this may well explain why you've been unable to find the connection.

It was just a general question, but triggered by a 32cM Ancestry match and her English grandmother who emigrated with her sister in 1880 - the year their father died. The match has two "brick walls" with her English gt-grandparents. The family are in London in 1861 (with two daughters: Sarah Jane and Sarah Ann) and Bristol in 1871 with three children who all emigrated - one is listed as Sarah but was registered as Eliza and seems to have used various names in later life! The gt-grandmother has Wiltshire as the place of her birth in 1861 (Bristol in 1871). To complete the known information, Sarah Ann died in Bristol 1865 and a son (1871-1875) was born and died in Bristol and, like his father, was buried in a non-conformist plot in Greenbank Cemetery, Bristol.

Shared matches fall into three groups - descendants of the sisters who emigrated, descendants of "Huggins" from the Grittleton/Hullavington area of Wiltshire (including myself and "cousins") and a few "not confirmed" (usually with no or minimal tree). Most (all?) descendant matches of the grandmother and her sister have latched onto "Bristol" and selected a local as the gt-grandmother though this person is married and living in Gloucestershire in 1861! I cannot possibly relate their suggestions for the gt-grandfather, who may be of Irish descent!

 

[peter]

While the 'shared match' feature at Ancestry provides some of the useful information chromosome browser tools might provide, its big limitation is that it only includes 'close cousin' matches.

I believe, but don't know for certain, that the primary reason for limiting shared matches is to help users avoid making assumptions which aren't justified by the data (a big problem for those who use other sites but don't understand conf0unding factors such as endogamy, pile-up areas, and crossovers). Given sufficient data there are any number of ways to jump to false conclusions.

I used to use all these tools - I didn't manage to knock down a single 'brick wall' with them. Hundreds of hours of my time were wasted. I put it down to experience - experience that others can benefit from if they choose.

 

[Pauline]

I didn't manage to knock down a single 'brick wall' with them

I wouldn’t necessarily expect to break down any brick walls, but these tools (if used carefully) can help with sorting out in which part of your tree a match might belong, and that in turn can help towards breaking down brick walls.

 

[peter]

I wouldn’t necessarily expect to break down any brick walls, but these tools (if used carefully) can help with sorting out in which part of your tree a match might belong, and that in turn can help towards breaking down brick walls.

That's the theory - but it didn't work for me despite all the time and effort I put in, and the multiple tests that I manage. Has it worked for you?

 

[Pauline]

Has it worked for you?

Yes and no! Not having any close family who’ve tested (apart from sister), the shared match feature is the best chance I have of sorting out which matches belong to which grandparent, where that is not apparent from trees. I have broken down one of my more significant brick walls but I still have more than half of my ‘4th cousins’ matches that I can’t relate to a particular grandparent.

 

[Charles]

Yes and no! Not having any close family who’ve tested (apart from sister), the shared match feature is the best chance I have of sorting out which matches belong to which grandparent, where that is not apparent from trees. I have broken down one of my more significant brick walls but I still have more than half of my ‘4th cousins’ matches that I can’t relate to a particular grandparent.

Agree with Pauline - shared DNA cousins are useful most of the time. I have found links to over 70 cousins though the clues in shared DNA (some in distant cousins) mainly by using methods from Peter's Masterclass. I also use my own methodology based on DNA circles and sharing other researchers matching lists. Most of my 4th cousins are now assigned to a group that is either a GG grandparent couple or a GGG grandparent couple. I have now found a paper trail to 48 out of 242 4th cousins and 95 of my distant cousins (some of these are through ThruLines but majority are through my research). However those searches are becoming more difficult so maybe Peter was further down his investigations than I was before the appearance of ThruLines and Shared DNA.