Genetic Genealogy has defined, through common use, a very specific term named triangulation. Triangulation is a way to weed out false-positive matches by verifying 3 or more autosomal DNA testers have a common matching segment that they all share when compared with all the other testers in the group. When such a case occurs, they are said to represent a triangulated group and almost assuredly share a common ancestor that contributed that triangulated segment. Note that pile-up regions can still cause a triangulated group that is still a false-positive match or possibly due to a very ancient, common ancestor. If and when GPS catches up, a triangulated group (not due to pile-ups) is a proof that a common ancestor must exist. Once you find a triangulated group, it still requires further genealogical work to find the common ancestor(s). (In most cases, it will be a common set of ancestral grandparents and you likely cannot tell which grandparent the triangulated segment came from. Proof of the common ancestor in atDNA testing requires not only finding the likely shared ancestor using knowledge of the DNA results (such as the number of shared matching segments) but also proving that it cannot be any other ancestor that they share. Such negative proof, not just the possible answer, is very arduous to develop and often requires a full tree and segment mapping to various ancestors. Often termed visual phasing. Arggh. When will the terminology stop!)
It is not enough to have common, shared matches to form a triangulation. This because the matching can be on different chromosomes and different segments. And thus from different parents contributions. Common, shared matches is a good start to looking for a clue to a common ancestor and triangulated group but is not a rigorous proof. Only when you have three or more DNA testers with a common, overlapping matching segment do you form a triangulation or triangulated group.
ISOGG has given a very different, very early meaning for triangulation in their glossary. One that is not the meaning in common use by tools and technologists since the advent of atDNA segment matching around 2012. Their defined meaning, that supposedly originated in 2006, just involves two DNA testers with matching results (likely was considering Y DNA only then) and talks of matching DNA results AND genealogy between only two testers (not three). In their definition, the third point on the triangle is the purported common ancestor that is not tested. In the common form of the term used by the industry, triangulation only involves atDNA testing and not genealogy. A minimum of three testers is needed to form a triangulated group. To distinguish these two concepts of triangulation, we will call the ISOGG definition trivial triangulation (or original triangulation) and the commonly used one (matching) segment triangulation (or triangulated group formation). If not qualified, we always mean segment triangulation on this website and virtually all industry websites which use the simple term triangulation.
Early indications by project admins here are that triangulated groups are a way to break through strong endogamy matching that clutters atDNA match lists for individuals in ethnic groups with strong endogamy. But triangulation does introduce the burden of requiring 3 (or more) in a common segment match; and not just a singular match between two testers.
To verify a triangulated group, one needs to, for each member of the group, verify ALL the other members have an overlapping matching segment in the same chromosome at the same location when analyzing all testers in a pairwise fashion.
Note that there is still a possibility, if only three testers in the group, for it to appear as a triangulated group when none exists (not just due to false matching segments like in a pile-up region). This can occur if each tester shares an ancestor through a different parent with each of the other testers and happens to share a segment at the same location on the same chromosome. The likelihood of this happening, even for three full siblings you try to triangulate, is minute. Diagrams will help here and be introduced soon, So like the minuscule possibility that two siblings could share no DNA, it is an extremely remote possibility that an apparent triangulated group does not share a common ancestor that all three share. But still needs to be mentioned.
It is not enough to have common, shared matches to form a triangulation. This because the matching can be on different chromosomes and different segments. And thus from different parents contributions. Common, shared matches is a good start to looking for a clue to a common ancestor and triangulated group but is not a rigorous proof. Only when you have three or more DNA testers with a common, overlapping matching segment do you form a triangulation or triangulated group.
ISOGG has given a very different, very early meaning for triangulation in their glossary. One that is not the meaning in common use by tools and technologists since the advent of atDNA segment matching around 2012. Their defined meaning, that supposedly originated in 2006, just involves two DNA testers with matching results (likely was considering Y DNA only then) and talks of matching DNA results AND genealogy between only two testers (not three). In their definition, the third point on the triangle is the purported common ancestor that is not tested. In the common form of the term used by the industry, triangulation only involves atDNA testing and not genealogy. A minimum of three testers is needed to form a triangulated group. To distinguish these two concepts of triangulation, we will call the ISOGG definition trivial triangulation (or original triangulation) and the commonly used one (matching) segment triangulation (or triangulated group formation). If not qualified, we always mean segment triangulation on this website and virtually all industry websites which use the simple term triangulation.
Early indications by project admins here are that triangulated groups are a way to break through strong endogamy matching that clutters atDNA match lists for individuals in ethnic groups with strong endogamy. But triangulation does introduce the burden of requiring 3 (or more) in a common segment match; and not just a singular match between two testers.
To verify a triangulated group, one needs to, for each member of the group, verify ALL the other members have an overlapping matching segment in the same chromosome at the same location when analyzing all testers in a pairwise fashion.
Note that there is still a possibility, if only three testers in the group, for it to appear as a triangulated group when none exists (not just due to false matching segments like in a pile-up region). This can occur if each tester shares an ancestor through a different parent with each of the other testers and happens to share a segment at the same location on the same chromosome. The likelihood of this happening, even for three full siblings you try to triangulate, is minute. Diagrams will help here and be introduced soon, So like the minuscule possibility that two siblings could share no DNA, it is an extremely remote possibility that an apparent triangulated group does not share a common ancestor that all three share. But still needs to be mentioned.
External References
- ISOGG Triangulation definition (not the same as given here)