Thanks, Tim, For those of us not using phasing techniques, and instead rely on various segment rules, I think the Triangulation technique is still an excellent tool. Using the utilities at 23andMe, GEDmatch and FTDNA, along with the closest cousins I can find (with known MRCAs), and working with every Match I can find with segment data, I now have about 80 percent of my atDNA segments assigned to a parent. When I say "my atDNA segments", I mean pseudo-segments which are a combination of segments from Matches. For example, if I can Triangulate A (21-34Mbp); B (22-45Mbp); C (25-36Mbp); D (27-52MBp) on one Chr, then I create a pseudo-segment (21-52Mbp) which is what I got from our Common Ancestor. In comparing A B C D to each other I usually find that their shared segments are cover more than the segment I got. This is OK - it just means our Common Ancestor started out with a segment that was, say (15-60Mbp) and different ones of us got different, but overlapping parts of it. Thanks again for providing some real- world insight on the question at hand. If I ever find such an anomaly, I'll be sure to report it. None yet. Jim - Sent from my iPhone - FaceTime! On Nov 21, 2013, at 1:36 AM, "Tim Janzen" <tjanzen@comcast.net> wrote: > Dear Jim, > In my case, the question is whether or not the 6.5 cM segment came > down through person #10 or person #15 on my mom's grandmother's pedigree > chart, an older version of which may be seen at > http://wc.rootsweb.ancestry.com/cgi-bin/igm.cgi?op=PED&db=janzen&id=I8. The > 6.5 cM segment could have come from a different ancestor than the rest of > the 21.7 cM segment. I don't personally get entangled in situations like > the one you raise because I am already segregating my parents matches on any > one segment of an isolated chromosome before I compare them to each other. > I do that by separating those matches into those who also match me as > compared to those who do not match me. This separates the matches into > those matching on one chromosome and those matching on the opposite > chromosome. Therefore in my parents' match lists I am never comparing A to > B as in your situation. In any case, your situation will occur at least on > rare occasion. I haven't seen that exact situation, however. > Sincerely, > Tim > > -----Original Message----- > From: autosomal-dna-bounces@rootsweb.com > [mailto:autosomal-dna-bounces@rootsweb.com] On Behalf Of Jim Bartlett > Sent: Wednesday, November 20, 2013 8:19 PM > To: autosomal-dna@rootsweb.com > Subject: Re: [AUTOSOMAL-DNA] Finding shared atDNA ancestry without phasing > > > Tim, > > So you have Mom has HIR with A(21.7) > Mom has HIR on same segment with B(6.5) > and A has an HIR on same segment with B > > This is classic genetic Triangulation. > > Is there any reason to believe there is more than one Common Ancestor for > the 3 of them? > > The possibility that we are studying is: > > Mom & A have an IBD HIR from Mom's Dad and A's Mom > Mom & B have an IBD HIR from Mom's Mom and B's Mom > A & B have an IBD HIR from A's Dad and B's Dad > All 3 IBD HIRs are on essentially the same location on the same > chromosome > number. > All 3 pairs would probably have different MRCAs. > > This is technically possible. The question is what is the probability. > I've > not seen it (that I recognize), yet. > > Jim Bartlett

Dear Jim, My technique doesn't involve phasing. It simply involves using the results from a parent/child pair to segregate the matches in any particular region of the genome. The technique is incredibly simple to implement if one has data from a parent/child pair. If you use this technique you don't have to worry about doing all the extra comparisons that one has to do with people like match A and match B as in the example you provided. In any case, I still have to worry about some of the matches being IBS, particularly for HIRs under 10 cMs in length. I also have about 81% of my mom's atDNA now assigned to a parent. It will be really hard to get this to 100% unless I can in some way retrieve my mom's mother's sample from the SMGF. I don't think you need to call segments you have linked to a shared ancestor "pseudo-segments". Those are true segments of DNA even if you haven't yet phased them. You may not have figured out the ancestor that the segments came from yet, but they are valid DNA segments that came from either your father or your mother. Like you, I am constantly trying to triangulate on matches who share overlapping segments of DNA with my parents. Sincerely, Tim -----Original Message----- From: autosomal-dna-bounces@rootsweb.com [mailto:autosomal-dna-bounces@rootsweb.com] On Behalf Of Jim Bartlett Sent: Thursday, November 21, 2013 4:55 AM To: autosomal-dna@rootsweb.com Subject: Re: [AUTOSOMAL-DNA] Finding shared atDNA ancestry without phasing Thanks, Tim, For those of us not using phasing techniques, and instead rely on various segment rules, I think the Triangulation technique is still an excellent tool. Using the utilities at 23andMe, GEDmatch and FTDNA, along with the closest cousins I can find (with known MRCAs), and working with every Match I can find with segment data, I now have about 80 percent of my atDNA segments assigned to a parent. When I say "my atDNA segments", I mean pseudo-segments which are a combination of segments from Matches. For example, if I can Triangulate A (21-34Mbp); B (22-45Mbp); C (25-36Mbp); D (27-52MBp) on one Chr, then I create a pseudo-segment (21-52Mbp) which is what I got from our Common Ancestor. In comparing A B C D to each other I usually find that their shared segments are cover more than the segment I got. This is OK - it just means our Common Ancestor started out with a segment that was, say (15-60Mbp) and different ones of us got different, but overlapping parts of it. Jim - Sent from my iPhone - FaceTime!