ICW (in common with) is not equal to triangulation!

+23 votes
1.0k views

I'd like to clarify the topic statement. 

In many sources on the internet the terms:

 

ICW (in common with)

and

Triangulation

are used in the same sense and are used equally. This is wrong!

 

ICW is a term that is especially used by FTDNA in their chromosome browser. Many people are assuming that because three people (let's name them A - which is yourself, B and C) are in-common-with each other that also automatically means they triangulate as well.

To clarify the term triangulation. It refers to a group of three individuals (they shouldn't be closely related like 1-2 generations) that are all matching at the same locus (location on a chromosome) with overlapping matching segments between them (they can start/end sometimes earlier or later than with others in the group but they must overlap). Triangulation is the only method to identify an ancestral segment that was given down from a common ancestor that all individual in this one triangulation group are descendants from.

What the ICW is only saying is that these individuals match each other. While you can see where you are (A) are matching B and also where you match C, there is no way you (alone) can see where B and C match each other (not on FTDNA or AncestryDNA - it's possible to see this on 23andme and on GEDmatch).

There is a likelihood that indeed all three of them match at the same locus but without the information been given by either B or C to you (A) where they match each other it's impossible to tell 100% sure.

So please contact B and C and ask them for this information and don't assume a triangulation!

I do have one comment from Doris Wheeler where she has a specific example where cousins from her maternal and paternal side are put into one ICW group by FTDNA but the resulting matching segments are on different loci (as proven by analysis on GEDmatch). Dr. Tim Janzen also stated that ICW is especially bad to use in "endogamous populations such as Mennonites or Jews because such groups generally share DNA segments with most other people with that ancestry."

 

I further refer to the following comment in this G2G thread by Darlene (http://www.wikitree.com/g2g/201219/dna-matches-how-do-you-spread-your-sheet?show=201700#c201700): "I recommend DNAGedcom.com in addition to Genome Mate. It helps you identify triangulated matches from Gedmatch and FTDNA"

I would like to clarify that based on the data that is available for downloading at FTDNA and AncestryDNA there is only one way to do triangulation, which is by downloading the raw DNA data from both sites to a tool like GEDmatch!

No tool in the world can work on insufficient data and do triangulation, as such the statement on DNAGedcom's page about the ADSA tool is wrong and misleading: "The Autosomal DNA Segment Analyzer (ADSA) is a tool that takes your data from Family Tree DNA or GEDMATCH and constructs tables that include match and segment information as well as a visual graph of overlapping segments, juxtaposed with a customized, color-coded In-Common-With (ICW) matrix that will permit you to triangulate matching segments without having to look in multiple spreadsheets or on different web pages."

I've asked Rob Warthen from DNAGedcom to change that misleading statement and clarify it. Again, it's not a fault of the ADSA tool, it's because FTDNA (and Ancestry) don't provide enough detailed information from their screens to do triangulation!

The only DTC DNA testing company that currently provides this information is 23andme. It also has the only tool to do triangulation with up to 5 people directly on their website (the Family Inheritance: Advanced tool).

tl;dr

Hope this G2G post clarifies that:

ICW <> triangulation

and that no tool can do triangulation without detailed data (meaning matching chromosome, with overlapping start and end position between all matching individuals).

in The Tree House by Andreas West G2G6 Mach 6 (61.1k points)
edited by Andreas West
Thanks for posting this, Andreas. Most of us who've done DNA testing have tested with only one company, so we don't have your insight into the differences in reports from different companies.
Possibly FTDNA did not have a chromosome browser at the time of the originial post, but they have had one for some months now.  It still is beneficial to export the data and import to GEDmatch, but it is possible to check were one matches on a segment and start/stop location.

Hi Thomas,

I'm regularly working with FTDNA and you're mistaken. They don't have a new "chromosome browser" tool, it's the same tool.

You still can't see where your two cousin Ernie and Bert match exactly (on which chromosome, start and end position). You can see where YOU match Ernie and where YOU match Bert. But that's not triangulation, it's a V (one part of the triangle is still missing.

tl;dr Nothing has changed at FTDNA and it most likely won't change because it's because of their privacy concerns that they don't allow us to see this information.

For a bit longer explanation feel free to read my extended FAQ that touches not only on this question but on many others regardings DNA and Genealogy: https://www.dnagenealogy.tools/extended-faq

I am new to FTDNA (and DNA in general) and had heard they did not even have a chromosome browser at one point.  So, we agree is that there is a chromosome browser in FTDNA that allows one cousin to see a match to other cousins including the chromosome number and start/stop.  We also agree that the transitive rule of equality is not applicable to matches.  If I understood your write-up, it is because they might be on different (maternal vs. paternal) chromosomes of a pair.  I did not know that FTDNA did phasing for cousins rather only for close relatives.  But your point is valid even if one were to exclude the phasing as an issue as all those wild card alphabet letters (a, c, g, t, and about a dozen others) cannot be assumed to be transitive so A matches B does not mean A equals B.  Regardless of how we get there, there is no application of the transitive rule.  I get that!

I would say, though, that one can still triangulate using the FTDNA browser as long as one has the cooperation of at least one of the other two cousins.  So, if we have A matches B and A matches C, either C matches B or B matches C would provide the triangulation.  If I missed something here, help me out.  This is all new to me.
Thomas, they can match or not match at all, you can see that in another FTDNA tool (their matrix tool).

Once you know that Ernie & Bert are indeed matching as well, chances are good that they will also triangulate. But in order to verify this, you need to contact either one and they have to tell you where they match each other, the part that FTDNA isn't showing to use via their tools.

Please keep in mind that it's not only matching on opposite sides (maternal vs paternal on same locus) but more likely it's on a different chromosome and start/end position.

Hope this is clearer now
It was clear already.  I'm am glad that you agree with me on the limits of the FTDNA chromosome browser!  It seems that the only difference is that your were saying that one cannot use the tool independently to determine triangulation and I was saying that a pair of matches can co-operate in using the tool to determine triangulation.  Same thing, different words.
Yes, Thomas, now I understand. It's all semantics I'd say. Surely they (Ernie and Bert) have to use the tool in order to tell you where they match.

That makes you dependant on one of the two responding to you or being still alive.

It could be much easier but FTDNA doesn't want us to know this without asking other people.

I think we're perfectly aligned. Thanks for your clarification.

1 Answer

+5 votes
Andreas, it is very wrong of you to state that I referred to ICW and triangulation as being one and the same thing.  I, in no uncertain terms, did not and have never referred to ICW and triangulation as being the same.  I would appreciate if you would remove my name from your post, as you are wrong in stating that I said that.

I am not a newbie with DNA.  I tested my father's mtDNA almost ten years ago and have had an ongoing interest in DNA ever since.  Shortly thereafter I tested his Y-DNA and then his auDNA almost two years ago.  I have been researching, analyzing and mapping my chromosomes ever since.

I work diligently to help others get interested in and understand DNA, making them aware that ICW (in common with) means that the profile person and another person share other people as matches, and that triangulation is when the profile person and two or more others share a common segment on a specific chromosome.

I will again reiterate that, if you're tested with FTDNA, you can upload your data to DNAGedcom.  You then can use their ADSA tool to help identify triangulated groups.

You are misleading people by your last statement about triangulation "...meaning matching chromosome, start and end position between all matching individuals."  People need to understand that all of your matches do NOT have to have the same start and end position to be part of the triangulated group.  For example, on my father's chromosome 2, one of his matches shares a segment with him from 222 to 234 (abbreviated for 222,182,992 to 234,603,627) (18.7cM); another match shares the segment from 222 to 240; another from 223 to 231; another from 226 to 235.  I've done a one to one comparison for each of them to one another, so I've confirmed it's on the same 'side' of the chromosome.  While they don't share the same start and end positions, they do in fact share a decent sized segment with my father and one another and form a triangulated group.
by Darlene Athey-Hill G2G6 Pilot (440k points)
edited by Darlene Athey-Hill

Dear Darlene,

I've changed the order of my statements to make it clear what you wrote in the other thread. Sorry for stating that you referred to ICW and triangulation as the same, I corrected that.

As for your comment on my last paragraph you probably have overlooked (as you mentioned that I mislead people) that I wrote earlier in my post about "that are all matching at the same locus (location on a chromosome) with overlapping matching segments between them". I marked it now as bold and changed the last sentence to include the word overlapping as well there, thanks for helping to improve my post!

You further wrote in your answer above: "I will again reiterate that, if you're tested with FTDNA, you can upload your raw data to DNAGedcom.  You then can use their ADSA tool to identify triangulated groups."

I once again went to the excellent documentation pages at DNAGedcom and tried to find where it says that you can upload your raw DNA data to DNAGedcom. I either wasn't able to find it or you can't. If I have overlooked it then please point me to the specific area where it's described as the only reference I could find is the following two parts:

"DNAgedcom's tools (such as ADSA) work by reading and displaying match data that you download from the testing company's web site (FamilyTreeDNA.com)"

and further down the same page

"This process will act as your agent to logon to Family Tree DNA behind the scenes and collect your match, segment and ICW data." (Source: https://www.dnagedcom.com/adsa/FTDNALoader.html.php)

 

As you note it clearly speaks about:

- your match

- segment

- ICW data

 

FTDNA doesn't provide you with matching data of your matches (aka how they match each other, the B vs C example in my original post) - despite to what DNA Genealogists are asking FTDNA for years to provide them with.

From the above it's also clear (unless you point me to another reference in the documentation on DNAGedcom) that you're not downloading (and subsequently upload) your raw DNA data to DNAGedcom. Even if you would do so, it doesn't help in doing triangulation unless all your corresponding matches are downloading their raw DNA data as well (same principle is true for GEDmatch, you can only compare at their site to people who downloaded and then uploaded their raw DNA data to GEDmatch).

I hope this is enough evidence that I will continue to stick to my statement that you can't do triangulation at DNAGedcom. You can access the top 200 triangulated matches from GEDmatch via their paid account (tier 1) and upload them to DNAGedcom but like I said, that isn't doing triangulation (this still happens at GEDmatch and I think the people behind GEDmatch deserve credit for doing that hard work). It's only visualizing triangulated data, albeit in a nice way (credit to Rob and others at DNAGedcom).

Please do also note that nowhere in my post I mention that you are new to DNA Genealogy.

Related questions

+8 votes
1 answer
+5 votes
2 answers
+10 votes
2 answers
+10 votes
12 answers
1.1k views asked Jan 19, 2019 in The Tree House by SJ Baty G2G Astronaut (1.1m points)
+13 votes
4 answers
+20 votes
8 answers
+8 votes
6 answers

WikiTree  ~  About  ~  Help Help  ~  Search Person Search  ~  Surname:

disclaimer - terms - copyright

...