Double Match Triangulator (DMT)

Recently, I received an e-mail from Louis Kessler, a Microsoft platform developer and genetic genealogist who has created a new tool, the Double Match Triangulator:

Double Match Triangulator (DMT) uses two FamilyTreeDNA Chromosome Browser Results files from two different people and combines them in an Excel file to produce a map of all the Double Matches and Triangulated segments between the two people and all third people. It provides a different way of looking at matching segments than the Chromosome Browser and helps identify Triangulation groups.

The program runs on Windows, is free for everyone, includes a short but comprehensive help file, and is available at: www.beholdgenealogy.com/dmt

Please note that this tool is only for Family Tree DNA chromosome browser download files and only runs on a PC.  There is no MAC version.

I’m already providing this functionality for the people whose kits I manage and have access to their match list, simply by utilizing a regular spreadsheet. In fact, I combine multiple people in a single spreadsheet to see who matches whom on the same segments. I’ll be demonstrating this in a future blog article.

However, I also realize that not everyone wants to do that manually, so this is a great tool for those who don’t, and it does the math between two people so you don’t have to. The only down side is that it only compares two match files at a time, but given Louis’s format, I don’t really know how he would incorporate more than two files in one comparison file. Besides, I’m not about to look a gift horse in the mouth and I think the DMT tool that Louis has provided is a great service to the genetic genealogy community.

Installation and Instructions

First, let me tell you that you must install this program on your PC. Now also let me say that I’m EXTREMELY leery of installing anything on my PC, especially from a relatively unknown source. However, I could tell from reading the documentation and information that Louis was indeed a genetic genealogist, and I decided the potential risk was worth the potential reward, so I did the install which went very quickly and easily. Besides that, curiosity got the best of me.  Yes, I’m apparently part cat, although that has never shown on any DNA test:)

Secondly, let me also say that BECAUSE this program is installed on your own system, and does NOT require you to upload match information into a public location, I’m OK with this. I would have serious ethical reservations if this program were offered on a public site where you uploaded matches to a public location, because your matches have not given permission for that to happen and it’s outside of Family Tree DNA where your matches expect their information to remain. However, since you’re only downloading to your own PC, for kits you manage or who have already given you access, I’m fine with this.

Third, don’t even think of trying to do this without utilizing the help file. I have dual monitors, so I opened the help file on one screen and the program itself on the other screen. The help file contains all of the instructions and column definitions, and you can’t run the program or make sense of the results without this information.

The Program

Below is the program screen where you select the two chromosome browser files that you have downloaded from Family Tree DNA to compare. Of course, these must be from kits that you manage or have access to in order to download the chromosome browser matches. Otherwise, you can ask your matches to e-mail you their match file.

dmt program

The Help button that holds instructions is at the right side below the option boxes, just above the line between grey and white.

The Help screen looks like this and is where you will find all instructions for how to select files, run the program and interpret the results.

dmt instructions

Scroll down to the bottom of this main help screen to see the various page selections under the title “The contents of Double Match Triangulator.” You will need to click on each one and read what Louis has written.

The contents of Double Match Triangulator

Double Match Triangulator Main Window
Excel Output File – Map Page
Excel Output File – Person Page
Log Files
Interpreting Results

The Results

After following Louis’s installation instructions, I downloaded three chromosome browser files that I manage and am familiar with so that I can determine if the program is functioning as expected.

I selected my cousin Cheryl, cousin Bill Lentz and cousin Rex Miller. All three of these individuals descend from the same lineage and they do match. I already know they have triangulated matches – although the common ancestor of their common matches is not always identified.

I would suggest creating a folder called something like “DMT Experiment” and copy the chromosome browser files you wish to use to that location. Your DMT results file will be in the same location.

After clicking “Run,” the program did not automatically pop up the completed match file, so I sat and waited before thinking to check the folder where I had put the files to use.

Two files were generated, Cheryl to Bill Lentz and Cheryl to Rex Miller. I’m going to use the Cheryl to Bill file for purposes of demonstration simply because it’s smaller.

dmt Cheryl to Bill

On the first (default) spreadsheet tab, labeled Map, we see the details of the Person A, Cheryl, matching to person B, Bill, with their common segment matches to a third person listed in the third column, Name-C. The next columns are documented in the Help file, but the column you are most interested in is the Status column which shows the “Full Triangulation” cells highlighted in green. These are the people where person A (Cheryl) matches person B (Bill) who both match person C, on the same segment(s).

dmt Cheryl to Bill status

While this matching “per segment” information is useful and interesting, Louis fortunately added a second tab, called “People” which shows one row per triangulated person. On the above spreadsheet, an individual match can be listed many times, for as many segments as qualify to be matching or triangulated. For example, you can see above that Donald is listed 3 times in the first few rows of this spreadsheet.

The People page, on the other hand, provides a summary with one line per matching person. So, on the People page, Donald will only appear one time.

DMT people

The People tab provides a significant amount of information that you’ll need to refer to the instructions to decipher.

However, the most important columns are the first six which show you the names of the two people whose matches are being compared, their common matches (Name-C) and summary information about those matches. Louis’s program did not color code these. I did, for purposes of discussion.

DMT People 2

In this case, the first match is between Cheryl F and Bill Lentz to Donald, who is Cheryl’s brother. They all three match on a total of 9 segments for a total of 47.02 cM. That’s significant.

The next two individuals, jerdon and james, also colored green, match on a significant size segment and jerdon triangulates on two separate segments.

These would both be considered strong triangulated matches.

The individual colored yellow is still a relatively strong triangulated match, especially in light of the triangulation, and in particular if the common ancestor is known.

Generally, when matching alone, meaning NOT triangulating, I would dismiss the smaller red segments as probably irrelevant. However, in light of the triangulation, and common known ancestors, I am not nearly so quick to dismiss these. I talked about small segment guidelines meaning how to and how not to utilize small segments in this article.

Among these small segment matches are known cousin Rex Miller, known cousin Roland, mother and me. I would not discard these, but I also would never “call” any relationship on one small segment alone. The good news is that when matching to others, on my DNA Master Spreadsheet, some smaller segments do form triangulation clusters. There is safety in numbers, whether utilizing smaller or larger segments. Of course, larger segments are always safest because they are less likely to be identical by population (or chance) instead of identical by descent. You can review the concepts of identical by descent, chance and population in this article.

Louis has included one more feature. The last three rows are people who match two of three, but not all three people. While these people don’t all match each other on the same segment, they do match a pair and a third person from this same line might well match those two. In other words, in the bottom row, Betty Jean matches Cheryl and Bill, both, but Cheryl and Bill do not match each other on the same segment where Betty matches each of them. More research is needed to determine whether Betty Jean is from the Lentz or Miller lines, or not. It is, however, a good hint to follow.  If you would like to review the concept of match groups versus triangulation groups, you can do so here.

What’s Next?

For me, I need to sign in to Family Tree DNA to see who jerdon is. Jerdon is the second highest triangulated match and I don’t know anything about that person. I surely hope they have a tree online, and if not, I’ll be e-mailing them shortly. In the mean time, I’ll be doing ICW and trying to figure out which line jerdon is from by utilizing other available autosomal tools.

Have fun with the DMT tool, and thank you Louis.  You can catch up with Louis at his blog where he provides additional explanations for his tool and the underlying concepts.

29 thoughts on “Double Match Triangulator (DMT)

  1. Thank you for sharing. How is this better or different than/from Tier 1 at Gedmatch? It seems more hoop-jumping than Gedmatch.

    I also wish someone would devise a program to chart our ancestors on an X fan lineage. Doing it manually would take more time and patience than I have. In my next lifetime, I plan to come back with patience and very long legs. LOL

    • You don’t upload your match file to GedMatch. The only matches GedMatch finds are people who upload their files there. This takes all of your matches at FTDNA and compares them with another person.

  2. Thank you, Roberta, for your thorough review of DMT. I enjoyed reading your thoughts and ideas about it and was somewhat amused by how few double matches you have in your two test runs. Most of the files I tested with are Ashkenazi and we’re talking 8,000 double matches and 4,000 triangulations in each of them.

    A quick tip to find the output files is to press the “Folder a” button.

    The missing a-b matches may be quite useful and are a bit different than your match groups because your match groups are still single matches, not double matches. Just yesterday, I figured out where some of the missing a-b matches may be coming from and posted: “Triangulation and Missing a-b Segments” on my blog to make that information available: http://www.beholdgenealogy.com/blog/?p=1799

    I am hoping that you and other experts will figure out some step-by-step ways to use DMT that I’d be able to program into it, so that everyone can benefit.

    Louis

  3. I am going to try out DMT immediately. Thanks for bringing it to our attention.
    I have another serious concern as I am over 80 and would like to leave my tree to my children/grandchildren. You are young yet but must have thought about the problem from your platform of more experience. My tree is currently on Ancestry and at home on Ancestry’s family tree maker. With Ancestry making changes, I wonder what to do? Is there a tree maker which I can put my tree and can easily be given to a number of people without costing them? I would also like fo it to be my everyday tree. Maybe you would be willing to write an article on the subject. I am sure that many others would be grateful for it.
    Thank you

    • Hi Emory, I wish there was a good answer. I’m not sure there is. I used to say Rootsweb, but anymore Rootsweb is kind of going by the wayside and I’m concerned that Ancestry will eventually close it down. Geni merges everything into “one tree” and My Heritage isn’t free beyond 250 people in a tree, but they keep your tree if you drop out. So I don’t have a good answer. I wish their was one. Maybe WikiTree.

    • I think putting a tree with Familysearch.org is the best choice. It’s free and is most likely to be stored over time as it is run by the Church of Latter Day Saints whose dedication to preserving genealogical records the world over has been demonstrated for many decades, and made freely available to anyone without regard for religious affiliations — ie you don’t have to be a LDS member,

      • Diane Redfern, thank you so much for your advice which I had not thought of. I’ll look into it right away.

      • I have found you have to be very careful with familysearch. It is a great resource, but many trees and other information are guesses and not that well documented. Of course, may of us have some of that, but make the effort to document. Unfortunately many of the guesses seem to propagate, even if you delete them later. I always debate about only posting things that I have documented and not just based on family history. Maybe Roberta will have some guidelines there.

      • I have found you have to be very careful with … EVERY tree that ANY person or group EVER put together on ANY service or ANY website or in ANY book. Many of them are great resources, but many trees and other information are guesses and not that well documented. [Just to clarify wadavis’ comment 😉 ]

  4. Roberta, Do I understand correctly that you need to run the Chromosome Browser separately for each match you want to be able to triangulate so that you can put them in separate files? Could you run all of them (if less than 6) and then separate them by each individual or does that just mean you need to do a bunch of sorting first?

    • No, I think we have some confusion here. You pick two people and you download their entire chromosome browser match list, for each of the two individual people. You have to sign on to each account to do so. Louis’s program compares only two people, person A to person B and looks for common matches between the two people, specifically on the same segment. You do download a person’s entire match list from the chromosome browser page, but it’s not related to running the chromosome browser itself.

      • It is turning into an fantastic journey of new info, though sorting things out is a different issue. I did find that you can use that folder option and it will run all the chromosome files it finds against the “a” file. That saves a lot of time. We have been running it on 8 files total at this point and then putting them into a composite file. From the composite file you can sort and look for common names among all the files.

        The only problem I have had so far is finding any useful GEDCOM links to put folks together. In fact, one line has history back to England and Scotland, and now lives in Australia (never the USA). So how I connect is a pure mystery. On average, I have found the cM distance to the 3rd party (Missing a-b match) is about 40. We have been deleting the full triangulation matches that we typically already know about to focus on these others, since they provide a link that ICW does not provide. Thanks for bringing DMT to our attention.

  5. Thanks Roberta (and Louis). I hadn’t heard about DMT until seeing your blog. I’m a DNA addict already, so I have the spreadsheet – but I just ran it and I think this can be very useful in identifying crossover points when comparing matches with close relatives. It also helps make FT ICW more user-friendly.

  6. I have my uncle tested at FTDNA and his first cousin also tested there. I used your information from a previous blog to phase my uncle’s results against that first cousin. How would the Double Match Triangulator improve on that outcome?

  7. On GEDmatch, I have a couple of triangulated matches on chromosome 8 between about positions 7 and 14 and we all share an East Asian/Siberian/Amerindian segment that is identical on the admixture paintings to one another. Both of the other matches get 2-3% East Asian and Amerindian in their overall admixture results, whereas I get very little, but I do share this segment with them. Can anything be deduced?

  8. Pingback: Friday Finds 2 September 2016 – Copper Leaf Genealogy

  9. This seems like a nice program. I have only one issue, with one of the author’s comments in the notes about the program: “But don’t throw away the small triangulated segments: Person a may match Person c over 10 cM, and Person b may match Person c over 12 cM, but their overlap may only be 1 cM just because of the way their segment got passed down.” Well sure, but if they only overlap 1 cM, then they may as well overlap none at all, because the odds of that overlap being false are far too great. I don’t think you should draw ANY conclusions, even speculative, from a 1 cM overlap.

  10. Informative ideas , I am thankful for the information – Does anyone know where my business might get ahold of a blank MT Beneficiary Deed document to complete ?

  11. Roberta, Thank you so much for providing hints for the Triangulation app. I do have a question about how Chromosomes work. Even though I can see results in Excel which seem to put 3 or more people as matching on the same segments of a Chromosome, if the precise ancestor hasn’t been identified do I assume they are matching on one side or the other (maternal or paternal side) of that Chromosome?
    I have run the program on a known 1st cousin, my brother, and myself. It is great that the results also show other matches and where we have common segments. But when other people match us the same segments how much can I assume if some of those people may potentially match our family on both maternal and paternal ancestries?

  12. Heads up, the latest release does not produce a people page in the Excel output file as described in your post. I’ve sent a note to the developer.

  13. Pingback: 2016 Genetic Genealogy Retrospective | DNAeXplained – Genetic Genealogy

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s