Genetic Affairs AutoCluster Analysis

Like much of the genealogy community, I have been excited about the new tool at Genetic Affairs called AutoCluster Analysis.  Blaine Bettinger made a very nice overview of the tool Forming and Using Genetic Networks with Genetic Affairs, found at YouTube.

After the tool was introduced to the genealogy community, the ability to do AutoCluster analysis at Ancestry was disabled.  Thankfully that feature is up again, as of late December, and a user can generate a report with matches at Ancestry, where many people find more matches than at other sites.

First consider whether you are comfortable with Genetic Affairs having your login information for the other companies.  I am, but consider it.  If so, sign up for an account at Genetic Affairs.  Add the DNA account information for all the accounts you might want to analyze.  Then choose one to process.

Before you process your first kit, consider the benefit of adding notes to your AncestryDNA matches.  Adding notes makes sorting through matches a lot easier.  I add notes to each of my matches with the names of the shared ancestor or couple, and the relationship.  It will make AutoCluster Analysis and any analysis in general a lot easier.  At least add notes to the matches where you have shared ancestor hints; Ancestry does the work for you (though not all of the hints are accurate all the time).

The cost of each AutoCluster analysis is 25 credits, where a credit is one cent.  Each person starts with 200 credits free. This should be enough to try out the tool and see if it is something you find useful.

Then start the AutoCluster analysis.  I use limits 600 cM down to 9 cM, for example.  Genetic Affairs will do the analysis and send an email.  Save the attached files.  Unzip at least the html file.

Open the html file.  A big grid will appear with your matches in groups.  This is my grid.

lj_autocluster_all

Each group, or cluster, has a color.  Matches of individuals between clusters is shown in gray.  Each group is a set of people where most of the people in that group match the other people in the group.  Much of the time, if you match a group of people and they match each other, you all share common ancestry.  That is what makes this analysis useful.  Some of your matches will have family information and some won’t.  The matches that do have information can help speak for the others in the group and place them in your tree.

Scroll down to see details about the matches.  Go through your clusters.  Each represents a part of your tree.  Of course, not all of your tree is represented, depending on who has tested.  If you used notes as suggested, it may be easier to see how each group is related to you.  A number of people in that group may have the same ancestral couple, or ancestors in the same line.

lj_autocluster_cluster5

This is an example of a cluster (with the names obscured).  There are seven people in the group.  The top two I have identified as descendants of my ancestors Georg Huebner and Caroline Wuest.  The other five are likely descended from this couple, or descended along one line of their ancestral couples.

The chart can be resized, for example so that other fields can be shortened to reveal more of the notes field.  You can click on a name and go right to the match page at Ancestry.  You can also click on info about a tree and go right to the tree, but I find that less useful.  I like to go to the match page first and see the tree displayed with any common surnames pointed out.

Overall I would definitely recommend checking out this tool.  It can help place matches who haven’t posted any family info into their part of the tree.  I may help reveal groups of related individuals you may not have noticed before and lead to more ancestor discoveries.

 

Posted in DNA

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s