Rob Warthen of DNAGedcom announced the Collins’ Leeds Method 3D yesterday. Once you run a chart, what’s next?
Above is a 3D chart I ran sorted by cM, not inclusions. Below is a chart I’m using to label the clusters. It starts with the 4 grandparents of the test taker: Stewart (A), Merrill (B), Peters (C), and Coppenbarger (D). These are in standard order from paternal grandfather on the top to maternal grandmother on the bottom.
Next, I added ahnentafel numbers. The “home person” is 1, their father is 2, their mother is 3. As you continue, each person’s dad is double the number of that person (so, #2’s dad is #4) and their mother is double plus one (so, #2’s mom is #5).
I added a column to the left of my 3D chart where I wrote a letter/number combination for each person whose relationship I knew. This label shows are LIKELY common ancestors. In the first cluster, these are almost all D2/D3 or a question mark if unknown. This means this cluster of cousins LIKELY all descend from J.R. Coppenbarger and Elizabeth Bennett. (Neat!)
In the second cluster, the top two people are labeled C2/C3 while the rest are labeled C6/C7. This group is LIKELY all descended from Carl Peters and Guntherine Werther, C2 and C3. But, more specifically, most of the group is descended from Johann Werther and Amalie Reinhardt, C6 and C7!
The next group are also C2 and C3, but they are LIKELY a different line than cluster #2.
The fourth cluster has B8/B9 and B16/B17 along with three unknowns. B8 and B9 would be the parent of James Merrill. B16 and B17 would be his paternal grandparents. So, this is LIKELY showing a very specific Merrill ancestry line.
(Isn’t this exciting!?!)
This labeling system also gives a good visual of what the grey squares represent. You can see that it is often the overlap between, for example, two different “C” groups. So, they are likely from the same grandparent line! In this particular example, grandparent C and D are from the same small town, so there is sometimes overlap just because of other marriages between the same families in this community.
One other comment… this labeling system also helps with HALF relationships as you can put ONLY the label, for example C6, of the common ancestor.
I’d love some feedback! Please give it a try and let me know what you think!
I haven’t yet followed all of your ahnentafel logic/connections, but yes this is the concept I’ve been using for the last 2 years with my spreadsheet matrices, and illustrates the paternal/maternal split and beyond that I saw in my bio parent search and have posted about in various forums.
From an adoptee standpoint (where there is no luxury of having a familial history or ahnentafel early on in the process), this is a goldmine of information and much easier than manually creating a spreadsheet and pivot table.
Using the ahnentafel as the basis of the familial groupings is also a great ordering method. I also add to that order with an extra A/B/1/2 label that I found handy for lists, etc. With my spreadsheets, since early on I was only able to “see” two obvious groupings I labeled one Group A and the other Group B. As I identified more relationships and clusters I took a page from the Y and mtDNA trees, and named the grand parent clusters A1, A2, B1, B2 etc. Grandparents then become A1a, A1b, A2a, and A2b, etc. Granted the sorted ordering comes out the same as in an ahnentafel but it visually helped me associate the clusters without using another “cheat sheet” to equate the ahnentafel numbers to familial lines.
In my theory (and yours ?) the gray “connecting” people should also give insight and corroboration to the grandparent and great grandparent familial “splits”.
Finally I’ll add that this is why I also suggest always sorting by cM, plotting the unclustered matches, and doing one or multiple runs which go from highest cM match to lowest cM match.. so you see ALL outliers and cluster connections.
When doing multiple cM runs to get the complete picture, you need to make sure you overlap cM values so that you will capture all outliers (grays) that connect both “upstream” and “downstream”.
Hi, Jim. Thanks for your wonderful comments! I would love to see an example of how you do this with adoptees.
Yes, I believe they gray “connecting” people can give wonderful insight into the connections between these groupings. I like sorting by cM, too, but I would also love to understand more about running multiple runs to see “ALL outliers and cluster connections.”
Thanks again! I’ll be reaching out to you. 🙂
Dana
Wow, I just tried this method since it appeared on dnagedcom recently. Words do not do this justice how easy it is to sort matches to their corresponding line. I’ve been looking for something like this to help sort out matches to find my 2nd great biological grandfather’s line. It makes it so easy with the html version having links to available ancestry trees. This seems to be promising as I’m going down the blocks even now as I write this message.
Jason, That’s terrific! I am so excited about this new method. I hope we see many brick walls fall because of what we will be able to “see” with these charts!
When I tried to run the new tool in DNA Gedcom it took a couple of hours and then said there was no ICW file. I have only used DNA Gedcom for importing to Genome Mate Pro from Gedmatch and Ftdna. Am I supposed to run something else first?
Hi, Patsy. I highly recommend asking technical questions at the Facebook group, DNAGedcom User group. There are a lot of techy people who understand that end of it a LOT better than I do. 🙂 The only other suggestion I’d have is to make sure you have gathered the ICW before you try to run the charts.
Hi Dana,
I love this tool! I am having trouble downloading the Collins’Leeds 3D Chart to my computer so I can add columns and label the matches. I did Ctl A, Ctl C and then selected keep source formatting on my Excel spreadsheet but I am not getting the cluster graphics. Any suggestions?
Hi, Carol. I see you are getting help on the Facebook group. For anyone else, two Facebook groups that can always help are Genetic Genealogy Tips & Techniques and the group DNAGedcom User group. There are great people on both sites who can answer questions!
I had the same comment at first but I then used dnagedcom client to download all my matches, trees and icw matches again (as I had done so the first time before the new feature appeared). After I did so, it worked just fine.
Now my only problem is trying to print out the chart. When I do, the chart prints all except the colored clusters. Bizarre. I’m not sure if it is some kind of printer issue or something else. If I find out, I will let you know.
P.S. I too would appreciate info on using this in adoptee searches.
Please disregard my prior comment regarding printing the clusters etc. Was just user (me) error. Works just fine if you are using the excel file rather than the html file. (But you need to do some formatting to print grid lines etc)
Hi, I been trying to do a CLM for 20 to 350 cM from Ancestry for several days now. It seems to be stuck on creating cluster GEDCOM 120/228. II have tried stopping and starting again, but it keeps getting stuck at 120. What can I do? Should I keep waiting.
Hi, Jan. If you’re on Facebook, I suggest asking the question in the “dnagedcom user group.” Both Rob & John are great at quickly answering any questions and solving problems.
Looking for tutorials, like this one, that are in German or other languages. Not finding any that can be used to help people, such as war babies, in finding their fathers from England and the US, etc. Can anybody translate and create companion pages in other languages?
Google translate can translate html pages. Just google “google translate”