Step 7: Pedigree Triangulation
Start by clustering your matches into four groups using the Leeds Method. There will likely be one cluster for for each grandparent. There are several tools that can help with this task. They are covered in our Intro and Applied Classes.
Use all second to fourth cousin matches (at least) and include more distant matches that also have trees.
You should end up with a group that looks like this picture.
Consider sorting the clusters to more easily see which matches go together.
The clusters are based on match triangles. Members of a match triangle most likely share a common ancestor.
There are three people in the blue match triangle below: you, your closest match, and a shared match.
Note that Closest Match and Shared Match are in the same cluster.
Compare the trees of the matches in the cluster. If your matches don’t have trees, you may have to build their tree for them.
Find the Most Recent Common Ancestors (MRCA) among the trees in that cluster of matches.
Tools that aid in finding the common ancestor, also known as Pedigree Triangulation, are covered in our Intro and Applied classes.
- Confirms the cluster and MRCA
- Checks for NPEs
Consider the predicted relationship range to guide how far back in the tree to look for the MRCA. Note that some third-party tools including DNA Painter and the Combined Relationship Prediction Chart Tool may be more helpful than company predictions.
You may need to contact a match to help build your tree back. More about contacting matches including example messages can be found under the DNA Section / Contacting Matches Tab of the Resource Library.
Use relationship prediction to place yourself in part of the tree. Be aware you may fit a generation or two earlier or later than where your match’s fit in the tree.
The following diagram shows a few of the possibilities.
Optionally, you can Test your guess (hypothesis) using DNA Painter’s What are the Odds (WATO) tool ($)
Repeat for another cluster, starting with the nearest match in the next cluster. You may find it faster to work on the collateral lines of the other cluster.