Tim Thanks for the examples. Four thoughts for all: 1. Tim didn't mention if the Matches matched each other, on this segment. We are particularly looking for 3 people who all match each other on one segment, and each pair got their atDNA from a different ancestral line. This is very rare, and we're trying to see if anyone has documented this, or if it is so rare that we can dismiss the possibility. 2. As we get beyond close cousins, we'll find there are shared segments and shared ancestry, but the shared ancestry is not from the shared segments. I have many of these cases, including one yesterday: my maternal uncle, Nathan Baker, did Y-DNA 10 years ago and helped prove the "Gunsmith" BAKER line; last night's atDNA match is also a Y-DNA match on this BAKER line - so we are definitely 7th cousins on my maternal BAKERs. Except in my atDNA spreadsheet I've got several folks who match him, who also match my Dad (5-way Triangulation); and a bunch of other folks with close cousinship on my Mom's side, whom he does not match. It is clear to me that this new Baker match and I have another Common Ancestor on my Dad's side, per the atDNA segment... 3. With Colonial Virginia ancestry, I have several Matches with multiple Common Ancestors. Like Tim, I label them as ALTs until I can Triangulate on the genealogy side. 4. With a large segment (and usually closer cousinship), we can pretty much count on that segment splitting, somewhere, into two smaller branches, both ancestral to the first. This is rational as we get one large segment from a parent (covers the whole chromosome); which we find is split into a few large segments from each grandparent; and those are further split into segments from the great grandparents; etc. (these splits are the crossover points). Sometimes they are found in the next generation (going back into the ancestry), and sometimes the segment stays pretty much in tact for several generations (going back), before it is split. The limit on splitting, of course, is the 7cM threshold as a Match is not reported below that level (except at GEDmatch), so we never "see" the next split. Actually, the above discussion is from my perspective as a living descendant looking back. What really happens is the atDNA from our ancestors is passed down, and new segments are created ! from the parents of each ancestor through recombination. This is the split that we see looking back. And this is why it will take about 500 discrete segments, each assigned to an ancestral couple (usually), to map our 44 atDNA chromosomes. The ancestral couple will be split apart, going back, into a even smaller segments we may find at GEDmatch. Jim - Sent from my iPhone - FaceTime! On Nov 20, 2013, at 3:31 AM, "Tim Janzen" <tjanzen@comcast.net> wrote: > Dear Jim, > I have a situation I came across about 3 weeks ago that is similar > to what you are looking for. There is a segment on chromosome 7 that my mom > shares with her 2nd cousin. I thus know that this segment was passed to my > mom from her paternal grandmother Anna B. Armstrong (b. 1876). About 3 > months ago I got a match on this segment from someone who has Dewey Ancestry > on two ancestral lines. The HIR is 21.7 cMs long and contains 4074 SNPs, > which is unlikely to be IBS. This was nice since Anna B. Armstrong has > Dewey ancestry through Sally Dewey (b. 1781). This person is both a 5th > cousin and a 6th cousin through the Dewey line. I thus mapped this > particular segment on my mom's chromosome map to Sally Dewey. Then about 3 > weeks ago my mom got another match on this same segment from someone who has > Bigelow ancestry. This person is a 5th cousin once removed on the Bigelow > line. The HIR is shorter at 6.5 cMs but it contains 1398 SNPs and is also > found in my brother, which makes it unlikely to be IBS. Thus, the segment > in question could have come down through Russel Bigelow (b. 1763). Both > Russel Bigelow and Sally Dewey are ancestors of Anna B. Armstrong on > different ancestral lines. This certainly gave me pause since I am now not > certain exactly which ancestral line from which this particular segment was > passed to Anna B. Armstrong. I have decided to label this particular > segment as being from either Russel Bigelow or Sally Dewey until I get other > matches on this segment that help me better sort this situation out. There > are a number of possibilities here: > 1. My mom's match with Dewey ancestry has Bigelow (or another allied > ancestral surname) ancestry in their pedigree chart somewhere. > 2. My mom's match with Bigelow ancestry has Dewey (or another allied > ancestral surname) ancestry in their pedigree chart somewhere. > 3. The 6.5 cM portion of the segment came through the Bigelow line and the > match with Dewey ancestry also has Bigelow (or another allied ancestral > surname) ancestry while the other portion of the segment came through the > Dewey line. > > I think we will find more of these types of situations in the future as we > all do more chromosome mapping. > Sincerely, > Tim Janzen >

Dear Jim, The match who shares the HIR that is 21.7 cMs long with my mom also appears as a match on the same segment of chromosome 7 in Family Inheritance: Advanced at 23andMe with the person who shares the 6.5 cM HIR with my mom. I don't have the raw data files for these two matches, but I believe it is highly probable that the HIRs are both IBD given that my brother also shares the same HIRs with both people. We need to move to phased data for these types of comparisons in the future. Hopefully 23andMe and FTDNA will eventually provide that for us. Hopefully Ancestry.com will eventually provide us with the HIR data and will improve the quality of their phased data. I think it is important to note that it is essentially impossible for two or more people to share a phased autosomal haplotype of any significant length and have them get it from different ancestral lines. The phased haplotype must have come from a shared ancestor at some point in the past. However, that shared ancestor isn't always apparent to the testees. 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 5:17 AM To: autosomal-dna@rootsweb.com Subject: Re: [AUTOSOMAL-DNA] Finding shared atDNA ancestry without phasing Tim Thanks for the examples. Four thoughts for all: 1. Tim didn't mention if the Matches matched each other, on this segment. We are particularly looking for 3 people who all match each other on one segment, and each pair got their atDNA from a different ancestral line. This is very rare, and we're trying to see if anyone has documented this, or if it is so rare that we can dismiss the possibility. 2. As we get beyond close cousins, we'll find there are shared segments and shared ancestry, but the shared ancestry is not from the shared segments. I have many of these cases, including one yesterday: my maternal uncle, Nathan Baker, did Y-DNA 10 years ago and helped prove the "Gunsmith" BAKER line; last night's atDNA match is also a Y-DNA match on this BAKER line - so we are definitely 7th cousins on my maternal BAKERs. Except in my atDNA spreadsheet I've got several folks who match him, who also match my Dad (5-way Triangulation); and a bunch of other folks with close cousinship on my Mom's side, whom he does not match. It is clear to me that this new Baker match and I have another Common Ancestor on my Dad's side, per the atDNA segment... Jim - Sent from my iPhone - FaceTime!