Patterns developing

This has been a slower month for family history blogging, but in the background the activity continues. I'm currently in the DNA Skills Workshop with YourDNAGuide. Earlier in the month I shared a list of research questions, but for Skills I've focused almost entirely on trying to breakthrough on my the families of my 3rd-great-grandmother Bridget and the mystery O'Briens who went to Pennsylvania.

Since starting with YDG two years ago, I've been trying to develop the knowledge and skills to work with DNA matches. I think I've come a long way, building clusters, and learning how to apply various tools. It has certainly been a challenge with an absence of matches to Bridget's line. I might have been looking at the matches in the wrong way, and without additional elements necessary to find the right patterns.

In July I started to notice clusters of Australian matches on my Dad's results on MyHeritage. While some of these people have Irish ancestry and connections to other branches on my Dad's maternal line, it has been tough to distinguish whether a few of them trace back to the O'Brien side or Bridget's side. I tried a new approach where I used matches in common between my Dad and O'Brien descendant EWM on MyHeritage. He shares 228 cM, and as of earlier in the month, they had 210 matches in common. I created a spreadsheet, and listed the name of every match, amount of shared cMs, amount of longest block, and other notes such as country or the name of the branch where I suspected they fit on my tree. Essentially I was creating and separating the network of matches.

When that exercise was done, there were clear groupings:

• Matches also in common with the Pennsylvania O'Briens
• Lee, Lambert, and Brown-Welsh descendants, not connected to the lines of John O'Brien and Bridget
• Mystery cluster in Australia/New Zealand and Suriname
• A group connected to a descendant of the Hanleys of La Salle County, Illinois, pointing back to Roscommon, Ireland
• An Irish group with links to Roscommon/Longford, Ireland

I was able to separate 85 people from the 210 who seem to be O'Brien/Bridget and worth exploring further to see how they are related to each other. Then I went back to my groups of matches on Ancestry, checking these MyHeritage matches to see who showed up in both places, and repeated the steps with FamilyTreeDNA and 23&Me.

In the meantime, this month Ancestry introduced in ProTools the ability to filter matches by ancestral region. Using this feature on my Dad's matches and the groups I had already created for O'Brien/Lamon and O'Brien/Bridget, with Leinster, Ireland as the region focus, I was able to see matches who had Roscommon and Longford connections. There were a few matches that I had not placed in the earlier groups, including one match with 44 cM. Looking at this person's shared matches of matches, I could then see a cluster of Australians and some of these people were also on MyHeritage, the earlier chart I created from EWM's matches, and at 23&Me. This feels like a breakthrough.

Two years ago I had an absence of matches pointing to Bridget's side of the tree, now I have a growing spreadsheet of very real leads across the different DNA testing platforms. In some cases, for those who tested at MyHeritage, FTDNA, 23&Me, or loaded kits to Gedmatch, I can see chromosome segments in common with me and my Dad's results. This has also helped narrow in on potential Bridget matches.

It makes sense that some descendants of Bridget's family lines left Central Ireland for better opportunity across the ocean in Australia, New Zealand and America. My next step is to see how these people match to each other, build out some floating trees, and see who has lines connecting back to Roscommon and Longford.