Hi Jim Thanks for your comments. I think I am good with the true matches now but the false positives can these only be determined by comparing the raw DNA when no genealogy match is found and is there a program that can do this? Karen On Sat, Oct 19, 2013 at 10:12 PM, Jim Bartlett <jim4bartletts@verizon.net>wrote: > Karen > > In general I agree with Tim. I would add that Triangulation has a very > specific requirement: A=B=C=A. That is each pair (AB,AC,BC) must have a > shared segment (usually at least 7cM, but perhaps smaller with some risk of > a false positive); AND each of these shared segments are significantly > overlapping - such that all three - ABC - have, say, at least the same 7cM. > You can accept a little less overlap, by accepting a little more risk that > it's not right. With so much randomness in atDNA, we cannot guarantee > anything. For example the segments themselves are not precise - the end > points are often fuzzy. But it works a high percentage of the time. > > If you are A, and you match B and C on basically the same segment; you > still need to determine if B and C share that same segment. This is fairly > straightforward using tools at 23andMe or at GEDmatch. At FTDNA the ICW > tool indicates if B and C have a shared segment somewhere. Often they share > the same place you do, but sometimes they don't. I looked at a lot of these > cases a year ago, and - from memory - about 2/3 ICW were on the same > segment and 1/3 was not. Or something like that - it was high enough, for > me, to investigate more, but not high enough to conclude there was > Triangulation - my judgement call. > So I usually ask B and C if they match on the same segment we share. > Either one can confirm it. This is a simple question for them, and if they > are on the same segment with you, this is very powerful information for > them, too. Particularly as the Triangulated groups grow beyond 3 folks to 5 > to 10 to 20... Such large groups should have enough researchers in them > willing to work on place/time matching - clearly if a Common Ancestor > doesn't become evident, we are dealing with a brick wall for many. > > When you have Triangulation, you know the shared segment (among ABC) is on > one chromosome, and that the 3 of you will have a Common Ancestor. > (Assuming you are one of the ABC). > > Now if any two in ABC have a known Common Ancestor, there is a good > probability that all three of you have the same Common Ancestor. When > dealing with deep Colonial roots (and similar populations) the possibility > exists that you and a Match could have more than one Common Ancestor. So > it's always wise to find at least one more Match who agrees on the > genealogy for this segment. This means waiting for additional Matches - but > don't worry, they are coming... > > Jim - Sent from my iPhone - FaceTime! > > On Oct 19, 2013, at 2:19 AM, "Tim Janzen" <tjanzen@comcast.net> wrote: > > > Dear Karen, > > The triangulation feature is a method that quickly tells you which people > > are also matching a known relative or a genetic cousin whose > relationship to > > you is unknown. You need to be careful how you interpret the results, > > particularly when you are dealing with endogamous populations. At this > > point I would suggest that you not jump to the conclusion that A and B > are > > related to you through the ggg grandfather that you share with your > cousin. > > Perhaps they are related through that line and perhaps not. I think you > > need data from known first, second, and/or third cousins to help you sort > > issues like this out. You need to use their data to map your dad's (and > > your chromosomes) so you can have some idea where each of these segments > > came from. You can then look at the matches on any one segment and can > have > > better insight as to whether or not they could have come from a specific > > ancestral line, such as through the ggg grandfather that you share with > your > > cousin. At this point, all I would do if I were you is to enter the data > > from A and B on your match list and wait for data to accumulate from > other > > matches at 23andMe, Family Finder, and GEDmatch to help you sort out > where > > the genealogical connection is on the segments you share with A and B. > > Sincerely, > > Tim Janzen > > > > -----Original Message----- > > From: autosomal-dna-bounces@rootsweb.com > > [mailto:autosomal-dna-bounces@rootsweb.com] On Behalf Of Karen Hodges > > Sent: Friday, October 18, 2013 1:42 AM > > To: autosomal-dna@rootsweb.com > > Subject: Re: [AUTOSOMAL-DNA] Family finder matches what to make of them > > > > Hi Tim > > > > My cousin and I match on Chromosome 4. This is my first DNA match besides > > Dad so I haven't yet needed to map anything.Dad should match my 4th half > > cousin as it is on his maternal side of the family tree but doesn't. I > > share a 3G Grandfather with my cousin. I descend from the first wife and > > they descend from the second. I have contacted Family tree DNA to check > > that an error has not occurred and it is currently being investigated. I > > don't see a genealogy connection to my Mum's tree. > > > > Match A matches me on chromosome 1 and Match B on chromosome 22 both are > > listed as 5th to remote cousins. Dad also matches A in the same location > as > > I do on chromosome 1 but does not match B although he does have another > > match to a person with the same surname. My cousin and I both show match > A > > and B when we triangulate each others name but we don't match them in the > > same locations. <snip> > > > > ______________________________ > For answers to Frequently Asked Questions about mailing lists, please see: > http://dgmweb.net/MailingListFAQs.html > > > ------------------------------- > To unsubscribe from the list, please send an email to > AUTOSOMAL-DNA-request@rootsweb.com with the word 'unsubscribe' without > the quotes in the subject and the body of the message >

Karen False positives from FF ICW can be determined by asking the two Matches in question if they match each other on the same segment you match them. Either one can confirm it. Also you can easily compare them at GEDmatch to see if and how much and where they share (if they'll upload to GEDmatch). Also you can confirm them, one at a time, if they match a close cousin (your close cousin can confirm the segment if you don't manage their kit). I would never think about comparing raw data. If they won't cooperate at all (as above), they surely won't give you their raw data, either. Jim - Sent from my iPhone - FaceTime! On Oct 19, 2013, at 7:52 AM, Karen Hodges <rowantreek@gmail.com> wrote: > Hi Jim > > Thanks for your comments. I think I am good with the true matches now but > the false positives can these only be determined by comparing the raw DNA > when no genealogy match is found and is there a program that can do this? > > Karen > > > On Sat, Oct 19, 2013 at 10:12 PM, Jim Bartlett <jim4bartletts@verizon.net>wrote: > >> Karen >> >> In general I agree with Tim. I would add that Triangulation has a very >> specific requirement: A=B=C=A. That is each pair (AB,AC,BC) must have a >> shared segment (usually at least 7cM, but perhaps smaller with some risk of >> a false positive); AND each of these shared segments are significantly >> overlapping - such that all three - ABC - have, say, at least the same 7cM. >> You can accept a little less overlap, by accepting a little more risk that >> it's not right. With so much randomness in atDNA, we cannot guarantee >> anything.

Dear Karen, I agree with Jim Barlett's comments on triangulation. I have one other technique I find helpful, particularly for HIRs that are relatively short (7-10 cMs) for which I have no other people matching one of my parents on the same segment. I also compare my two brothers and me against the person in question in 23andMe, FF, or GEDmatch. If the person matches one of my parents and also matches one of my brothers or me at the same location then the HIR is highly likely to be IBD. If it doesn't match one of my brothers or me then the HIR is more likely to be IBS. There is only a 12.5% probability that any one DNA segment that my parents have will not have been inherited by one of my brothers or me. Tim Janzen -----Original Message----- From: autosomal-dna-bounces@rootsweb.com [mailto:autosomal-dna-bounces@rootsweb.com] On Behalf Of Karen Hodges Sent: Saturday, October 19, 2013 4:53 AM To: autosomal-dna@rootsweb.com Subject: Re: [AUTOSOMAL-DNA] Family finder matches what to make of them Hi Jim Thanks for your comments. I think I am good with the true matches now but the false positives can these only be determined by comparing the raw DNA when no genealogy match is found and is there a program that can do this? Karen On Sat, Oct 19, 2013 at 10:12 PM, Jim Bartlett <jim4bartletts@verizon.net>wrote: > Karen > > In general I agree with Tim. I would add that Triangulation has a very > specific requirement: A=B=C=A. That is each pair (AB,AC,BC) must have a > shared segment (usually at least 7cM, but perhaps smaller with some risk of > a false positive); AND each of these shared segments are significantly > overlapping - such that all three - ABC - have, say, at least the same 7cM.