The previous article in this series is Introduction To Using GEDmatch Part 2 (The One-To-Many Comparison Tool).
It’s taken me awhile to post another article in this series because I was hoping the Triangulation tool would be reinstated by now. Since it has not, let’s explore the People who match one or both of 2 kits tool that has recently been reinstated. For each of the matches we get to our kit using the ‘One-to-many’ matches tool, it is very helpful to know who else happens to match both of our DNA tests. This can help us begin to identify shared segments between at least three people, forming the basis for properly triangulating the segment back to a common ancestor, or set of ancestors.
Clicking on the People who match one or both of 2 kits link will bring up the People who match both kits, or 1 of 2 kits page. Enter your kit number in the first box, and the kit number of one of your matches in the second box, leaving the other three boxes as they are, and then click the Display Results button at the bottom.
Clicking the Display Results button will take you to a page that looks similar to the following screenshot (I have greatly shortened the lists in the second and third boxes to make it easier to view an example of the results page). It will take a bit of time (up to a few minutes in some cases) for this results screen to come up. If you have waited at least five minutes, and the results screen has still not come up, hit the back button on your browser, re-enter the two kit numbers and click the Display Results button again. If the results screen still does not come up a second time, the site is most likely busy and it is best to try again later.
Your results page will contain three boxes. The top box is the one we want to focus on. The kit numbers listed in this box match both of the kit numbers you entered on the previous page. These are very likely to be relatives, but you need to verify this by running a ‘One-to-one’ comparison between each of these kit numbers and the two kit numbers you entered on the previous page. Check the two ‘One-to-one’ comparison results to make sure the same segment(s) and chromosome(s) are shared between all three people. If so, there is a common ancestor, or set of ancestors, shared by all three people.
Contacting each of the people with shared segment(s) and chromosome(s), and comparing your trees is the next step to identifying who is the common ancestor, or set of ancestors. Even though you know there is common ancestry because of the shared segment(s) and chromosome(s), not everyone’s trees will extend far enough back to identify the connection. This is why using Genome Mate, or a spreadsheet, is essential to keeping track of your comparison data to keep referring back to as you continue to add new comparison results.
[A note of caution] For some unknown reason, using the People who match one or both of 2 kits tool does not always produce a complete list of the kit numbers that match both of the two kits you are comparing against. If you get a message saying “No usable match found between these 2 kits. Is batch processing complete?”, do not automatically accept this as being correct. While it is indeed possible that the batch processing on one or both kits may not be complete, there is also the possibility that the GEDmatch tool is simply not returning a match list. In the next post in this series, I will show how you can use a spreadsheet program (such as Microsoft Excel) to run your own comparison to verify the results of the GEDmatch tool, and/or provide you with a results list when the GEDmatch tool does not.
If you get stuck or experience any problems while following the steps above, or notice that a screen has changed significantly from what I’ve shown in the screenshots, please let me know by commenting on this thread or sending me an email. Thanks for reading.
The next article in this series is Introduction To Using GEDmatch Part 4 (Alternate Way To Find People Who Match One Or Both Of 2 Kits).
This article last updated 7 Oct 2014
Adventures In Genealogy Research: No Stone Unturned/The Wright Stuff 
Thank you!! This makes sense now. I knew about the “people who match one or both” but didn’t think about the next step of doing one to one on each of them. This brings it all together.
Thanks for the feedback, Sarlert. I’m glad this was helpful, and I appreciate you letting me know. Feel free to make suggestions for future topics you’d like to see covered.
Pingback: Introduction To Using GEDmatch Part 4 (Alternate Way To Find People Who Match One Or Both Of 2 Kits) | Adventures In Genealogy Research: No Stone Unturned/The Wright Stuff
Is doing the process you describe above basically just the manual version of the hopefully soon to be back segment triangulation tool?
After doing the people who match one or both, I did a one-to-one between each match and myself. There are several that match me on the same chromosome/segment, but when I do the one-to-one compare between the matches they don’t match each other on the same chromosome segment that they match me on. So what does that mean? I am related to those matches, but not in the same way they are related to each other? I don’t quite understand how two matches match me on the same chromosome segment, but don’t match each other on that same segment, unless one is matching my paternal and one is matching my maternal chromosome.
Thanks for this series of articles. It’s one the most helpful out there about Gedmatch.
Hi Brian,
Thanks for your questions. You are correct. The steps I’ve been outlining are indeed a manual way to perform the same function that the GEDmatch triangulation tool provides. Like you, I also hope the triangulation tool returns soon, but this method I’m showing allows one to continue working with GEDmatch data without the triangulation tool being available.
In regards to some of your matches sharing the same chromosome/segment with you, but not with each other, I have come across this same situation with my own data. While it is possible that one is matching your paternal side, and the other is matching your maternal side, it is indeed a bit puzzling that the segments and chromosomes would be exactly the same. Your questions are reminding me that I need to more fully investigate this situation to find out exactly what is happening. If you feel comfortable doing so, I would greatly appreciate you sending me an email with the specifics of your matches in this situation. I want to send your information and my information to the administrators of GEDmatch for review, in the hopes they can provide more specifics as to what exactly we are seeing, and/or if there may be some kind of error in the comparison data.
Thank you,
Dan
Hi Brian,
I want to apologize for taking so long to get back to you on this, and thank you again for allowing me to review this. There weren’t any inconsistencies found by the GEDmatch staff in the information you emailed me, so I’ve also gone through these to check them. Here is what I found:
Your initial analysis is correct. These matches are from opposite sides of your family tree. Whichever side the known relative is on (maternal or paternal), the person who shares that same segment with you, but does not match your known relative, will share a common ancestor with you on the opposite side of your family tree.
While I did not find this with the examples you’ve listed, something to be aware of is that GEDmatch sometimes won’t show a match on a one to one comparison, despite the match appearing on a one to many match list. This is because of a difference in the threshold levels for matches between the one to many tool and the one to one tool. The one to one tool threshold level is higher. If a one to one comparison initially doesn’t show a match between two kits, it is a good idea to check the match again using a threshold level of 3 for Centimorgans (cMs) and 300 for SNPs. The threshold for a genealogically relevant match is generally 7 cM and 500 SNP, but by lowering the threshold level and running the one to one match again, I have come across several matches which are just under either the default cM threshold level or the default SNP threshold level, and are close enough to warrant looking into further.
In regards to some of the segments appearing to be exactly the same, I have learned that this can more commonly appear to be the case due to rounding used by the test companies. There are set start and stop boundaries that are utilized, although these specific boundary points are different for each testing company, and a segment which ends just under a boundary point, and a segment which starts just above that same boundary point, will both appear to meet exactly on that boundary point. Instead, there is actually a small gap between them. If we could see the true start and stop points of the segments, it would make it more obvious that a segment on the paternal side is actually not starting (and/or ending) in the exact same spot as the segment on the maternal side.
I’m hoping this explanation make sense, and is of some help, but if not, don’t hesitate to let me know.
Best regards, and thanks again for your question,
Dan