Goran Runfeldt, a fellow genetic genealogist, has developed a killer app. You’ve heard of “The Terminator?” Well, meet “The Triangulator.”
Goran developed the Family Finder Segment Triangulator tool to run, using a user script or browser extension, on the Family Tree DNA site, after you sign in to your personal page. So there is no downloading, no spreadsheets, nothing messy.
The Triangulator tool is still in beta, so while the documentation is rather sparse, the tool is extremely intuitive if you understand triangulation.
What is Triangulation?
If you don’t understand triangulation, what it is, how it differs from match groups, and why you would want to utilize triangulation, may I please suggest that you read the following articles before utilizing the tool.
Concepts – Why Genetic Genealogy and Triangulation?
Concepts – Match Groups and Triangulation
Triangulation for Autosomal DNA
In a nutshell, triangulation provides you with a tool to show that not only do person A and B match you, on the same segment, but that they also match each other.
This means that they are not matching you on the same segment number from opposite sides of your family, meaning one person matching you from your mother’s side, and one from your father’s side. If they match other, as well as you, that means that they both descend from the same side of your tree (assuming they are not both matching you identically by chance.)
Family Tree DNA shows you, utilizing the chromosome browser, that two people match you, and on the same segment, but they don’t (yet) inform you about triangulation, although they are working on a triangulation tool.
In the following example, we have 5 known relatives to Barbara, whose background chromosome is black. As you can see, there are three possible triangulation points where at least two of the people match Barbara.
Just to be sure, I downloaded these matches to a spreadsheet to illustrate that these matches are not trivial in size – meaning based on their size, they certainly should be legitimate matches.
All three matching areas on this chromosome (grey, gold and blue) are large enough to be considered substantial, and when compared to the charts created by Philip Gammon in the Match-Maker-Breaker article, we see that there is almost no likelihood that these are false matches, or matches by chance. In that article, when phasing matches to parents, we demonstrated that 97% of the matches of 12cM or more and/or SNP density of 2800 or more phase to one or the other parent, meaning they are legitimate matches. At 15cM, 100% of a child’s matches also match a parent, except for the X chromosome.
All of these cousins descend from Barbara’s paternal side, from the same family line, so the chances are pretty good that they do all triangulate, but let’s see.
Installing the Triangulator
First, you’ll need to install the triangulator.
My choice is to utilize the tool in Chrome, as I had difficulties with Internet Explorer compatibility. Chrome works just fine.
Goran has provided installation instructions for various browsers here.
If you’re installing this tool in Chrome, be sure to sign in to the Chrome web store while using Chrome to install the free app, or the store will ask you to download Chrome.
The installation is super easy – just one click, literally.
Ok, now the hardest part is over and we can get busy triangulating right away.
Sign in to your account at Family Tree DNA, using the browser where you just installed the tool.
Click on your Family Finder matches.
You’ll notice something new right away, a new icon that says “dnagen tools” at the top of your Family Finder matches. That’s the Triangulator.
On your match list, select the people you want to triangulate, just like you were selecting the people to compare in the chromosome browser.
Your comparison list will be built, like always, on the lower left hand side of your screen.
To triangulate, instead of clicking on the Chromosome Browser button, you’re going to click on the new dnagentools icon.
You’ll see a little dropdown box that says “Triangulator.”
Just click on “Triangulator.”
You’ll see the progress bar as the tool calculates the relationships of the people you are triangulating to each other.
When the tool finishes, it switches to the Triangulated Segment tab, which is what everyone wants to see first, but you can always click on the Relationships tab to view the various relationships of the people you selected to each other.
All of the genetically estimated relationship of all of the people you’ve triangulated to every other person in the group are displayed.
When the Triangulator is finished, you’ll see the “Triangulated Segments,” tab displayed, assuming some segments do triangulate, with a small image of the chromosome beneath each triangulated segment. The area where the segments match to you is colored in orange and where the segments all triangulate is colored in red.
Additionally, the tool shows you the actual overlap range, the number of matching positions and the overlapping number of SNPs as well.
If you think you’ve died and gone to triangulation heaven, you have.
In order for you to easily transfer this information to your spreadsheets where you are triangulating your segments (you are, aren’t you???) and assigning segments to ancestors, Goran has provided a nifty tool for that too.
At the bottom, Goran has included downloads of:
- All matching segments for these people
- The triangulated segments for these people over the match threshold selected, which defaults to 5, same as the chromosome browser
- The relationships of these people to each other
Yes, you can lower the threshold, but just remember that as you do, the chances of the segments being identical by chance increases.
The Answer to Our Problem – Triangulation is Critical
In case you’ve gotten all excited about triangulating and forgotten that we were in the middle of a story problem, let’s look at our answer.
If you recall, there were three candidate regions for triangulating between Barbara’s known cousins on chromosome 3.
However, the Triangulator only shows two triangulating segments, the first and third. That means that the second of these large segments does NOT triangulate. That means that one of these third cousins matches Barbara on that segment in one of these three ways:
- By chance
- Because the overlapping matching region is too small to be considered a match
- One person matches from Barbara’s mother’s side and one from her father’s side – as unlikely as that seems with third cousins.
The most likely reason for non-triangulation is the third reason, given those large matching segment segment sizes.
While the first and third (grey and blue) segment match groups both triangulate, the middle (gold) region does not.
If you’re shocked, just remember that no matter how intuitive a match seems, and no matter how “sure” you are that two people from the same line of your family certainly must triangulate because they both match you on the same segment, without triangulation, you REALLY DON’T KNOW!
And you all know about assume, right? Been there, done that, got educated!
Triangulate removes the assume from the equation.
In this case, triangulation tells me that I need to look on Barbara’s mother’s side for a second common ancestor with either C. Lentz or W. Lentz.
Just so you know, I was suspicious of this result, but given that I have access directly to the kits of both C. and W. Lentz, because I tested them both, I verified that they don’t match each other on this segment, both at Family Tree DNA and at GedMatch. So this is no mistake.
This triangulation tool is a “goodness of heart” free application shared with the genetic genealogy community, and while Goran is willing to share, he doesn’t really want his inbox to be swamped. In the tool, he provides the following support information.
Goran follows the ISOGG Facebook group, so posting questions there will provide answers for you, and maybe for someone else following along too.
What if I Haven’t Tested at Family Tree DNA?
The Triangulator tool requires chromosome segment data, thankfully provided by Family Tree DNA. Therefore, this tool is not available for use with Ancestry data at Ancestry. You can, however, download your Ancestry DNA file to Family Tree DNA. Not everyone who tests at each vendor uploads to other places, so be sure to fish in all of the ponds, one way or another.
You can read about which vendors’ files are compatible to transfer to Family Tree DNA (and other places too) in the article Autosomal DNA Transfers – Which Companies Accept Which Tests?
The following chart shows transfer Files Accepted at Family Tree DNA.
|Vendor||Fully Compatible Version||Partially Compatible Version||Incompatible Version|
|Ancestry||V1 – until May 2016||V2 – after May 2016 to present|
|23andMe||V3 – until Nov. 2013||V4 – Nov. 2013 – Aug. 2017||V5 – Aug. 2017 to present|
Keep in mind that the current V5 version of the 23andMe test is not compatible at all at Family Tree DNA. The 23andMe V4 version, in use between November of 2013 and August of 2017 is only partially compatible, as is the Ancestry V2 version in use since May 2016.
If you upload partially compatible versions, you’ll receive your closest (meaning largest) matches, generally about 20-25 % of your matches that you would receive if you tested on the Family Tree DNA platform. However, you’ll be missing most of your matches, and you never know where that match you desperately need is hiding.
Note that this isn’t an artificial restriction imposed by Family Tree DNA, it’s a function of the other vendor’s chips only being partially compatible with the DNA processing chip used by Family Tree DNA.
If you want to see all of your matches and all of your segments, purchase the Family Finder test at Family Tree DNA.
A really big thank you to Goran and the user interface developer, Jonas, for this wonderful tool.
I receive a small contribution when you click on some of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.
Thank you so much.
DNA Purchases and Free Transfers
- Family Tree DNA
- MyHeritage DNA only
- MyHeritage DNA plus Health
- MyHeritage FREE DNA file upload
- 23andMe Ancestry
- 23andMe Ancestry Plus Health
- Legacy Tree Genealogists for genealogy research
What a timely and helpful blog post — thank you! A cousin and I are working to find our mystery MRCA and were struggling to completely understand triangulation. I checked my email, and here was your post. Reading the linked posts now.
Thank you for sharing the triangulator tool information. This will prove interesting. You always keep the community uptodate, and I appreciate it since I need all the help I can get.
Wow…that is a fun tool. And it definitely brought up more questions. My husband was adopted. We have narrowed his birth father’s surname to Lake, with Y-DNA testing. We have a friend who lives in the same city where my husband was born, with the Lake surname. The Family Finder test shows this man to be either an uncle or a first cousin to my husband. I was thrilled to find that. I would love to narrow that down more. If he’s an uncle, that means his only brother (deceased) was the father. If he’s a first cousin, one of Mr. Lake’s father’s brothers was my husband’s father. They are also all deceased and there were a bunch of them.
I pulled up all my husband’s matches with Lake as a surname in the surname list. When I ran the triangulator, none triangulated to my husband and the Lake we know is related, even when I lower the cM to 2. I ran it again, first bringing up all their common matches (121), then narrowing it to those with the Lake surname. I came up with only one match. They do triangulate, and all 3 are in the area of 4th or 5th cousin. Of course, not all their matches have a list of surnames, so it is difficult.
Am I doing it right? Is it safe to assume that those with the surname Lake in my husband’s matches, are NOT related to Mr. Lake that we know is a close relative, except for one?
Now I have another quandary. My husband and I do not show any match to each other, but we have a common cousin match. She is in both our match lists. I know where she is in my family, but don’t have a clue about where she is in his. It’s the kind of mystery that keeps me awake at night.
Searching the Google Chrome store does not find The Triangulator. Can you please provide a direct URL for the download?
Yes, click on the link in the article.
It’s called the “DNA Genealogy Experiment Loader”.
Cool. I just installed the triangulator and I think it’s alright. Gedmatch has a very similar utility called “People who match one or both of two kits”, from which you can also triangulate the 23 chromosomes at once. I got surprised to know that all the 27 ppl who match my mom and I share with us the same spot on chromosome 4, exactly the same big chunck there. I learned from you that it was given us by the same ancestor(s), I just don’t know what else can I get from that information.
Thank you for sharing!
I don’t know if my other comment went through so I am repeating it below. When I attempted to submit my comment I was redirected to logon to WordPress logon. Seems my learning curve is never ending.
What an exciting tool! Thanks to Goran, Jonas, and to you, Roberta, for sharing your knowledge and expertise. Most of what I understand about DNA I’ve learned from your blog. I am forever grateful for your generosity. Sue Garley
If I understand correctly, this tool is finding overlapping segments among people who are ICW rather than doing true segment triangulation. Is that correct? It’s a reasonable workaround for most people, but users should be aware that it’s not doing true triangulation and that the results can be misleading, especially if you come from an endogamous population or are looking at a pileup region.
According to the developer, this is not the case and the triangulation occurs through an API.
Could you tell where we other developers can find the FTDNA’s documentation for the API? Or is it hidden for larger public?
That was my initial reaction too, how can it be Triangulation rather than just the ICW we can already do, albeit more slowly?
However, this does indeed look to be true triangulation.
Certainly the csv download of the triangulated segments contains actual segment data for the others in the selection used for the triangulator, which aren’t available from my log on and on a quick check of kits I can access, look accurate, being the chromosome browser data for each pairing selected, right down to 1cM segments.
Thanks for informing us of this tool and explaining it (I have a lot of playing around to do). But I would like to point out that although this IS a free tool, it does look like there is a way to make a donation; it appears that from the familytreedna.com website we can all make paypal donations to the developer. Small amounts would add up. I plan to do so. It would be a “group general fund contribution” to dnagen I believe. Maybe you could post a link to the place on the familytreedna.com website to make this easier for people to find/do.
Watched your presentation, come over well. But wish I could have been there, but it was excellent to have it seen live.
Allready received your email about the triangulation app.
Amazing, you are such a jewel.
This extension is a fabulous tool. It may still be considered beta but the author did a great job of testing before the launch. Everything I tried worked. Thanks for your review, Roberta!
Roberta: When you press one of the CSV buttons at the bottom, where do the files get downloaded to?
To your PC.
I can’t find them anywhere. Do the downloads actually work?
I tested one and it did but I’m traveling now and can’t retest.
Check your Downloads folder.
Couldn’t resist trying the new tool. I love it but noticed something that I’m not sure is a bug or a feature. I have two FTDNA matches that I already know I share a large (~34cM) segment with so I ran a triangulation that included them and some other people with a similar max segment length. That turned up two more people that also triangulate in the same area. The edges however are fuzzy and that seems to have resulted in the extension identifying 5 separate segments all basically in the same place. Do you think its supposed to work that way or should I report it somewhere?
Goran monitors the ISOGG Facebook list so I would post there.
If A, B, and C share the same segment with overlaps, the tool will use the starting and ending points where all three people share with each other. If you add person D, he or she may not share the entire segment that A, B and C share. So another triangulation segment will be formed for all four of those people. It will be yet another triangulation that will only include where A, B, C and D all share the same starting and end point together. You will end up with two separate segments on the same chromosome, each located within very close proximity from one another.
Thanks, WW. After running some more tests I kind of figured that out. I believe that this is different than Tier 1 Gedmatch triangulation (at least the last time I tried it) which defines the triangulated group as the smallest segment common to everyone that it includes. I like the FTDNA extension much better because I think it is in effect showing you cross-over points. I’m just not sure how to make use of that information yet.
By running groups that include a first cousin from both my maternal and paternal sides I believe that I have been able to clearly identify some segments as being definitely maternal or paternal. So far I have found only one small segment where both cousins triangulate on the same small segment so I assume that a segment matches the same side of which ever cousin triangulates with the others. Now if someone would just write a utility that made it easy to find everyone who matched those segments I’d be in heaven. Hopefully either DNAgedcom with learn to handle bigger files or FTDNA will give us a way to limit the size of DNA dumps.
You can download the spreadsheet and then sort in order to find that info. Then triangulate with those candidates.
Wow!! This looks to be an invaluable tool – a big thanks to Goran and Jonas, and also to Roberta for letting us know about it (and explaining it so well) 🙂
This is a fabulous tool! Thank you for sharing! It does appear to be using an API rather than just my ICW list, which is evident when I study the downloaded info on segments and triangulations. In any case, it is fascinating to see just exactly how big/small a given triangulated segment is with 5 or 6 people.
This is my second attempt to post. Anyone else having trouble with this tool? I’m using a Chrome browser and installed the tool successfully but when I attempt to use I get the message “No triangulated segments with the chosen segment length”. I tried it multiple times with all my closer relatives.
Goran monitors the ISOGG list so I would suggest posting there.
Roberta, I appreciate the detailed explanation of the instructions you gave, the website actually refers people to your blog for them. This does look quite interesting, I will have to try it out with some of my suspected triangulations. I’ve not focused a lot on FTDNA matches, as I could not be sure that they triangulated, but now I will give that another look.
For now, I’ll take it on faith that the science behind this method is solid, but at some point, it would be nice to be reassured by experts in genetic genealogy that it is indeed sound.
I’m not quite sure what you mean by “it is indeed sound.” Whether you mean this tool or triangulation.
I do believe I qualify as an expert.
Triangulation is indeed sound, no question there! The “how” of this tool is what I’d like to see examined and commented on by genetic genealogy experts, you are certainly one of them. You say that it works through an “API”, and that seems vague to me, can you comment on how you believe it operates to disclose true triangulation as opposed to an image of it?
Your previous blog posts on imputation seem to explain why MyHeritage.com figures me as 51% Iberian, when it is obvious from the pedigree trees of my matches that I am indeed French-Canadian, which coincides nicely with the fact that this adoptee was born in Quebec City.
Have you tried the tool?
Yes, I have, briefly. I just took my top five matches and ran them through it, and got a result. I’m on vacation now, and my main computer is back at home, I would really like to try it with some folks who I’ve suspected as a triangulation group.
While I deeply appreciate tools that make identifying triangulation possible, I still would like to be assured that there is a scientifically rational basis for believing in them from people who understand them better than I possibly could.
Roberta, this is the best thing ever. Thank you for telling us about it. I’ve been playing with it since I read your post.
The limit of people you can select with the check boxes in FTDNA is 30. FYI.
Elizabeth (or Roberta, if you know), Could you explain what you are referring to about being able to select 30 people in FTDNA. I can only select 5 on the match page (and Roberta indicates this too). So I’m curious where 30 people can be selected with check boxes on FTDNA.
Sue, I had the exact same question, but now have the answer. Although FTDNA only allows 5 matches to be selected at one time, the Triangulator tool allows up to 30 … once it’s been run. Begin by selecting 2 – 5 matches and run the Triangulator tool. Once it has finished processing, close the tool. You’ll now be able to select up to 30 matches to serve as input to the Triangulator tool.
Yep. It’s like you have to finagle around with it and discover it accidentally. 🙂
I love the Triangulator and have been attempting to sort through my ever elusive Southerlands. Please look at the matches below and the position ranges for Segment 1 and Segment 2 — are they on different halves of Chromosome 5? Does this mean that M.C. Southerland matches Todd Turner on both his maternal and paternal line? Thanks Velma
Chromosome 5 (4 segments) Segment 1: 9.23 cM, 2,400 SNPs, 9,809,993 positions Members (3): Frank David Sutherland, M.C. Southerland, Todd Turner Position range: 109,792,316 – 119,602,309
Segment 2: 10.07 cM, 2,600 SNPs, 10,608,345 positions Members (3): M.C. Southerland, Todd Turner, Kathy Southerland Burner Position range: 109,398,887 – 120,007,232
Segment 3: 9.14 cM, 2,500 SNPs, 10,933,377 positions Members (3): M.C. Southerland, Todd Turner, Kathy Southerland Burner Position range: 124,608,738 – 135,542,115 Frank David Sutherland – Kathy Southerland Burner: No MatchFrank David Sutherland – Velma Joyce Southerland: No MatchFrank David Sutherland – Todd Turner: 4th Cousin – Remote CousinFrank David Sutherland – M.C. Southerland: 5th Cousin – Remote CousinKathy Southerland Burner – Velma Joyce Southerland: Full SiblingsKathy Southerland Burner – Todd Turner: 2nd Cousin – 4th CousinKathy Southerland Burner – M.C. Southerland: 1st Cousin, Aunt/ Uncle, Niece/ NephewVelma Joyce Southerland – Todd Turner: 4th Cousin – Remote CousinVelma Joyce Southerland – M.C. Southerland: 1st Cousin, Aunt/ Uncle, Niece/ NephewTodd Turner – M.C. Southerland: 2nd Cousin – 4th Cousin
| Roberta Estes posted: “Goran Runfeldt, a fellow genetic genealogist, has developed a killer app. You’ve heard of “The Terminator?” Well, meet “The Triangulator.”Goran developed the Family Finder Segment Triangulator tool to run, using a user script or browser extension, on ” | |
On the “About” tab it says “Please respect other people’s privacy and do not post any user identifiable information”. Roberta edited/deleted some details of the matches in the examples she gave.
Oh, dear. I’m glad I’m not in that list of names. 🙁
The “install” in Chrome as an extension did not work, the “bookmark” option does. Tried checking for people with “Ingalls” in their list of ancestors, seems to work OK. Known relatives are reported as relatives, etc.
I installed this yesterday for Firefox, and tried it just now. I got the message “Family Finder common matches API, required by this tool, is no longer available.” I’ll check to see if I followed directions, and also check Goran Runfeldt’s instruction page. I hope FTDNA has not done something to disable this tool.
I’ve been playing around with different strategies for how to select people to compare in the triangulator. I’ve found some interesting things but mostly it leaves me wishing I could then easily look for other people who might match on the same segment as the group that the triangulator identifies. Is there any utility on any site that can do this without a lot spreadsheet games?
Not that I know of.
Roberta, if I understand correctly what June is asking… wouldn’t Don Worth’s ADSA tool at DNAgedcom accomplish her needs? And I think maybe you did an article about the tool somewhere in your past blog posts?
Tom, I’ve tried playing with ADSA but I’m Ashkenazi and the build (which I think was through DNALand) can’t handle the file size. I’m almost up to 13000 matches and FTDNA does not currently provide a way to limit the output DNA file to anything other than the five matches you send to the chromosome browser. Otherwise it only lets you dump the whole file.
Ooops. Didn’t read carefully enough. Yes the build was through DNAgedcom not DNALand
Isn’t that what GEDmatch’s Tier 1 tool “Matching Segment Search” does?
No, it has been changed:(. Try it now.
Roberta, thanks so much for this very useful article. You say above, referencing your “Match-Maker-Breaker” article:
“In that article, when phasing matches to parents, we demonstrated that 97% of the matches of 12 cMs or more and/or SNP density of 2800 or more phase to one or the other parent, meaning they are legitimate matches. At 15cm, 100% of a child’s matches also match a parent, except for the X chromosome.”
For those of us without parents’ DNA to phase, can we too assume that any non-X-chromosome match of 15 cMs or greater would match one or the other of our parents and therefore be a legitimate (not IBC) match? Might any such match be IBS rather than IBD?
Thanks very much for continuing to educate so many of us.
Yes, I think that’s a legitimate assumption.
It’s probably the last of the Indian summer days for autumnal gardening but I am outside under a sunhat doing FTDNA with this “triangulator” all afternoon, thanks to you. Heavens. Loved your Ireland storm articles as well.
And BINGO! I just used this to pick up a triangulation which I hadn’t been aware of before, and I *think* it has put a huge hole in one of my longstanding brick walls. It connects up a lot of loose ends into a nice tidy bundle, and now I just need to go and test it out against everything else I can think of. If I’m right, then there are going to be some very happy cousins all over the place.
Roberta, thank you so much, you have no idea – well, maybe you do – how exciting this is!
Is it necessary to have at least one parent’s DNA to use this tool?
Thank you. Very easy to get the app on Chrome and easy to use.
I tried to install and it asked me to login or register. Since I am not registered, therefore no login, I tried to register and it told me it was not accepting anymore registrations now.
What asked you to log in or register?
Great tool. Much appreciated. Is there any way I can select more than 30 people (1 page) at a time to triangulate?
No. That’s the limit.
I’m a bit confused about this because I can only compare 5 at a time – certainly not 30! I use Firefox rather than Google Chrome; I wonder whether that’s why. Also, I’m finding that I get different results at different times. For example, I compared 5 people (let’s call them 1 to 5) and 1 and 2 matched me for 9.34 cm on chr 12 but when I compared 1 and 2 with 6, 7, and 8, 1 and 2 matched me for 22.5 cm on chr 12. Has anyone else had this problem?
The first time you run the tool, it’s five. The second time, without closing the tool and opening it again, you can select up to 30.
What I have done is select my 30, open that spreadsheet, then of those 30 that I selected, uncheck the ones that are not in the spreadsheet (because they didn’t triangulate with one of those 30, so I’m just going to ignore them), then go to the next page and keep checking, reviewing spreadsheet, and unchecking until I have triangulated enough to satisfy me.
You may not sleep for days!!!!
Not to spoil the party but how is this tool going to identify the third side of the triangle?
If you (A) match B and C, how does the tool tell you where exactly (which chromosome, start and end position) B is matching C?
This would be triangulation, to my knowledge and past unsuccessful tries with the FTDNA website this information isn’t revealed at all to person A (due to privacy concerns, as A could identify medical predispositions for B and/or C).
Roberta can you explain?
Or anyone else?
Otherwise this isn’t a triangulation tool, it’s only identifying two sides of the three that a triangle needs (hence its ICW – in common with, which doesn’t mean that B and C triangulate on the same locus).
Though I’d welcome it if FTDNA would finally give us access to the third missing triangle side! It’s not the tools, it’s what information is being accessible.
Medical issues could be revealed in the same segments as compared to you, so this isn’t any different in that respect.
Yes it is triangulation. Try it yourself on multiple kits you have access to.
I was also wondering about privacy issues. The Relations CSV and All Segments CSV downloads in particular seem to give more information on the relationship of the two other legs of the triangle than I perhaps have a right to know without their permission. “Relations” shows their estimated relationship, total shared cM, longest segment and total segments. “All Segments” displays the total segments they share with each other whether or not they share them with me.
Looks like there have been some changes that have addressed privacy. I used it today and only the Triangulator was available from the pull down menu as the tree functions have disappeared. With the results only the Triangulations CSV is available for download as the Results CSV and All Segments CSV buttons have disappeared. Listed relationships now follow the general categories of Immediate, Close, Distant and Speculative.
Yes, I wondered about that also. It was a lovely feature, but I wasn’t sure it ought to be there. However, even knowing that there is a triangulating segment – or a predicted match – already gives you information about the genetic relationship between two other people. I’m not sure how one decides where to draw the line.
Great addition to the site! Being able to look at 30 at a time is terrific – makes it so much easier to run comparisons as the database grows.
I scrolled through the replies to this post, so I hope I’m not asking something already asked. If so just say “scroll again.”
When I go to add the app to chrome I get a notice that adding the loader “can read and change my data on familytreedna.”
There was an ancestry.com chrome app that also had that warning. Does anyone have an idea what the nature of the potential changes to my data on FTDNA could be?
Never mind, I found some items on the internet. Google is the only one that ask for permission, but it’s the same for other browsers, but they just don’t “ask,” apparently. I should have searched it out in depth earlier.
Roberta, I’ve known your family (and related) match ours for a quite some time, but not how we are related. I found out today using The Triangulator that 12 of y’all that are stacked up on one chromosome with me are triangulated with me. And Dad has 4 segments triangulated among 14 matches on the same chromosome, including both your Ancestry kits. Very cool! Now I just have to get my tree branched properly to run into yours someday. In the meantime I’m going to toss in some of my close known cousins with those related to you that triangulated with my family to see if the picture gets clearer. Fun!
Very cool! However, I picked one of my toughest cases to start with, and I’m as confused as ever! I’ve compared 14 matches who all match me on overlapping segments of one chromosome (all over 7 cM, most over 10 cM, and some over 20 cM), and all of whom match at least one of the other 13 matches (usually more), and none of whom show any common ancestors in their trees. The Triangulator returned 41(!) segments on that chromosome, all of which overlap. And the matches don’t sort out into nice groups, but form something more like a vast web. What is going on?
OK, is anyone else having problems with FTDNA only letting you select 5 matches at a time? I was able to select more earlier today, but now it’s cutting me off at 5!
Run 5 the first time, then you’ll be able to select more.
Thanks! Not intuitively obvious!
Roberta, I was behind on my emails and just caught up to this article but I can’t seem to find the link for the Triangulator. Please tell me where it is. Thanks.
In the article under “Installing the Triangulator,” click where it says, “here” in blue.
I’ve noticed today that the Triangulator tool is not giving 2nd-3rd cousin, 5th cousin – remote designations and only Immediate, Close, Distant, Unknown. I don’t like this and there seems to a flaw in the app. My known lst cousin’s son is showing as unknown to me and my sister. I will report this to Mr. Goran. Something has changed in the last day–probably due to privacy issues again.
Someone in the Arsenault posted a note about this tool so I downloaded it and have given it a try but I’ve gotten a couple of strange results. I chose 4 matches so 5 triangulations; 1 of the matches does not triangulate with any of the other 4 (including me) yet when I look at the “Relations” tab, it says that this person is a close relation to myself and 2 of the other matches. I think I had the same thing happen with another set
OK so I think I’m getting the hang of this now and have figured out my own answer to my last post which just proves why triangulation is such a great tool 🙂 So I have three matches (two are siblings – let’s call them A, B and C) and all four of us match 8.79 cM on ch 1. C has a gedcom on FTDNA but the siblings do not though I have built a tree based on information they gave me and based on those two trees, I have determined they share five sets of common ancestors with the closest making them second cousins once removed. In another triangulation, a match I shall refer to as D matches myself, B and C for 12.64 cM on ch5. I have absolutely no information on match D other than that I share 158 cM with her since there is no tree and emails have gone unanswered; do these results suggest that D may be related to B and C through one parent while A is related through another?
I think I would need to see this drawn out to answer the question. To triangulate, you have to match all 3 people on the same segment, and they all also have to match each other on that same segment. I hope that helps.
Yes, I match persons A, B, C and myself all match for 8.79 cM on ch 1; myself, B and C all match for 12.64 cM on ch 5 (but not A who is the sibling of B). Hope that makes sense.
Siblings could have DNA from opposite parents on chromosome 5. Each chromosome segment needs its own triangulation.
Thank you so much, Roberta; I’m trying to get a good understanding of triagulation. I used this new Triangulator tool to compare myself, A, B, C and D. There were matches over many chromosomes but only 2 where four of us matched over the same segment. One was ch 4 (which I mistakenly identified as ch 1) where A, B, C and myself matched for 8.79 cM (but not D); the other was chromosome 5 where B, C, D and myself matched for 12.64 cM (but not A who is a sibling of B). So I’m wondering if this suggests that B may be related to C and D through one parent while A is related to C and D through another?
I logged in to my own kit as well as several of my others and this new ‘dnagen tools’ is not to be found so FTNDA has not set it up yet on all kits. Do you have any idea when all kits will have it available to use?
It’s an third party add-on. Read the article and follow the instructions for set up.
A late comment on something I just discovered. Some others may have already noticed it, but I don’t see it mentioned anywhere here, so FWIW…
Once you run the Triangulator once to get past the 5-selected limit, if you wish to run the Triangulator on a whole page (30) of already displayed matches, you don’t have to select all 30 check-boxes one by one, but just run the Triangulator with nothing selected and it will run on the 30 displayed. Makes things much easier for doing large mass runs. I had been hoping for a “select all” button, but this solves the problem. Saves a lot of time and clicks.
A cousin ( call him A) in my husband’s (B) list, shares 78 cMs, 48 of those on one chromosome and I suspect that one of his listed surnames might be the surname of my husband’s maternal grandfather’s father. ( we think the grandfather might have gone through his life by a different surname between b1875 and d1921, possibly because he was orphaned at an early age). Using this tool via the ICW for A, list, (about 50 names) I came out with about 25 names that match both A and B. However some are listed as speculative for A and distant for B and visa versa, and some are listed as speculative for both. Before I proceed with the match list, should the ‘speculative’ ones be weeded out for either or both A and B. Also the cousin and my husband are listed as distant by triangulator and 2nd to 4th by FTDNA.
Thank you for both your time and sharing the information about this tool.
I wouldn’t weed any out.
Thank you for your posts. I enjoy learning. I was wondering how the triangulator is different from the Matrix on FTDNA FF? The results seem to be the same.
The matrix tells you that people match each other. The triangulation tool tells you that it’s on the same segment.
The first time you run the tool, it’s five. The second time, without closing the tool and opening it again, you can select up to 30.
I finally found a way for it to do more than 5 but I have to filter my results in some way first and then it compares the entire page (ie. 30 plus me) which simply won’t work because it times out. The other thing is that I find I sometimes get different numbers when I compare just three at a time to when I compare 5; on one comparison it was 11 vs 14 which I considered significant. Consequently, I’ve now decided I’ll start with my top matches and create a separate spread sheet for each one; time consuming, I know but it makes it more manageable to compare results. I’ve emailed Goran Runfeldt about this but got no reply. Has anyone else had this problem?
I just ran the FTDNA triangulator tool with a different set of people this morning (2 known connections and one new one) and it’s finding NO MATCHES for any of them. is it broken? It did fine the first day I installed it (right after reading your block – thank you!).
This message is now found on the tool. “The team at FTDNA has closed what was perceived by a few users to be a potential privacy risk. Over the next few weeks the Dnagen.net team will be working closely with FTDNA to make sure that the triangulation data adheres to privacy standards. Meanwhile the Triangulator tool will not work. Sorry for the inconvenience!” I will blog when the tool is restored.
I finally found some time to install the Triangulator today – only to discover that FTDNA has suspended it 🙁 The FTDNA message says “The team at FTDNA has closed what was perceived by a few users to be a potential privacy risk. Over the next few weeks the Dnagen.net team will be working closely with FTDNA to make sure that the triangulation data adheres to privacy standards. Meanwhile the Triangulator tool will not work. Sorry for the inconvenience!” There’s no date to tell us when they suspended it, and no idea on how to find out when it’ll be restored – you have to go through the whole selection process and attempt an triangulation before you find out that you can’t. Roberta, is possible that you’d be able to give us a heads-up when it’s working again? Many thanks
It just happened earlier today and I have no idea for how long. Yes, I will print an announcement when it is restored.
Given how nicely it worked, it would be lovely if FTDNA could negotiate with Goran to adopt it and make it routinely available on the site. Maybe as an “opt-in” or an “unlock” or a “tier 1”, depending on their business model.
A good idea but terrible implementation. 1) Can’t get the tool to run under the Firefox on Linux, even though Greasemonkey seems to install OK. 2) Greasemonkey won’t work with anything but the Firefox beta – one can’t even run this on the stable version of firefox. 3) POut Chrome on and now FTDNA won’t let it work anyways at the moment because of privacy issues. Totally useless. Hope FTDNA releases their own Triangulator soon.
Gosh Tyler, I didn’t have any of those experiences and I installed it using three browsers. However, I too hope that FTDNA incorporates a triangulation tool into their toolbox.
No trouble at all installing it and using it with Safari (which is usually more problematic than Firefox). Useless for you, maybe, but fanstastic for others. I hope FTDNA resolves the privacy issues with Goran and restores the Triangulator, because given FTDNA’s track record implementing and upgrading features, I don’t trust them to come up with a decent tool themselves.
I had no trouble installing it via Firefox.
It’s not working right now, because FTDNA has shut it down due to possible privacy issues. Drat!
Fingers crossed that this can be worked out soon.
You may have mentioned it, but can one set the lower thresholds on the triangulation for cM and SNP’s to 4 cM and 100 respectively? I have an autosomal dna study (GEDMatch) involving distant 10 (5th 6th) documented American cousins (descendants of a mid 1750’s immigrant grandfather)) and a descendant of an Englishman who may have been (oral history and circumstantial evidence) the father of said American immigrant). I have found 3 triangulated segments (with the English descendant) thus far with two cousins and the potential English cousin on each segment. Segment sizes are as mentioned above. A 2nd new English descendant participant will be joining the study soon. Summary: Attempting to determine MRCA 9-10 generations back – a brick wall at the immigrants father to date.
That’s too lie to find anything meaningful. I don’t use anything below 500 SNPs or below 7 cM
Anyone know the status of the Triangulator? Hoping to see it back working again soon.
So do I!!!
I’ve almost given up hope of ever seeing that wonderful tool again. I know GEDMATCH has a VERY inexpensive one but, unfortunately, only a few of my major matches are on there. Not much sense in asking people to transfer their raw data either since response rates to my emails probably runs at a miserable ten percent or less. That’s my gripe for the day. If anyone can suggest a way to get FTDNA to revive it I’d be over the moon.
I’m working on it!!!