My First MyHeritage AutoCluster Analysis

This is a busy time for genetic genealogists. Companies seem to be competing with each other to get out new products. Genetic Affairs has been a leader in clustering analysis. They perform detailed clusterings for AncestryDNA, FTDNA and 23andMe. Their MyHeritage (MH) cluster analysis is a little different as it is done within MH. One set report is created and the detailed file of the segment information is not sent.

I got 100 people in the clusters. The thresholds were between 30 and 350 cM. The matching threshold for my matches matching each other was set at 20 cM.  These clusters were done in the old way in that the largest clusters were firs and the smallest last. There is no clustering of clusters. A first look at the AutoClustering shows that Clusters 2 and 3 match each other. There is also some affinity between Clusters 2 and 3 and Clusters 7, 9, 11 and 12. Colors now repeat every 12 clusters as opposed to the previous 10.

Identifying MY Clusters

I am familiar with many of the names already and have written Blogs about them. I’m going to first look at the low-hanging fruit and put them into a spreadsheet:

I only found two grandparent (GP) lines. I notice German and Russian names in Cluster 4 which is probably Rathfelder GP.

Splitting Apart Clusters 7 and 17

I have been working on Frazer DNA for a long time. As a result, I may be able to split Clusters 7 and 17, Note that I have the same common ancestor in Clusters 7 and 17. However, that is for Emily and Paul. Gladys has a different common ancestor.


This shows that Gladys has no known McMaster ancestor. This means that Cluster 17 should not include McMaster. I don’t think that I know that Cluster 7 is McMaster for sure, but it is more likely McMaster.

Clusters 6 and 16

Clusters 6 and 16 don’t separate as easily:

Ron, Stephen and my family share Clarke and Spratt ancestry. However, at a generation or so further back, we all share McMaster ancestry. It would take finding another common ancestor from someone in one of the clusters to further separate these out. Cluster 16 is interesting because the two other people who match Stephen have their ancestry in England. It is likely that Clarke and/or Spratt had English roots. McMaster was in Ireland for quite a time, so Cluster 16 is not likely a McMaster Cluster.

Red Cluster 1

I can identify Cluster 1 – at least at my GP level. One of the matches is JL. JL shows up when I do a ‘One to Many’ query on my paternally phased kit at Gedmatch:

This corresponds with my Hartley GP Line:

In the past, I have associated this match with my Massachusetts Colonial Heritage. This heritage is through my Snell side:

Here I added England in the notes for Cluster 16 from my previous section. I added Hartley in blue with information back to my Snell and Bradford 2nd great grandparents. Both these 2nd great-grandparents have Massachusetts Colonial ancestors.

I Need a Fourth Grandparent – Maternal Grandmother

Beth is the one I call my anchor DNA match. I have blogged about Beth here.

The other two matches in Cluster 15 have ancestors from England. It would take a bit of sleuthing and research to find the connections.

I now show all four grandparent clusters.

All I have to do is figure out the other clusters.

Leave a Reply

Your email address will not be published. Required fields are marked *