It is true that lots of 4cM IBD segments exist for all of us. But there are two problems: 1. However many there are, you would be swamped with many more IBS segments. 2. We don't have a good way to test for IBS, short of phasing which most folks can't do. Refer to my blog. All the segments in your body (on chromosomes in your DNA) are IBD! I call them ancestral segments. You only get your DNA from your ancestors - 1cM segments, 3cM segments, all of them. The segments we "see" in a chromosome browser or a spreadsheet are shared segments (with a Match), as determined by a computer program comparing your SNPs (from both your parents) with your Match's SNPs (from both of his/her parents). This shared segment is made up, fabricated. Over the short term - with only a few hundred SNPs - it's easy for the computer algorithm to "stitch together" apparently matching segments. As much as I tout triangulation, we don't really know if it works for very short segments. I'm comfortable predicting that triangulation works to sort out IBS segments over 7cM virtually all the time. These are typically longer segments, and I have yet to find any which don't work over 7cM. I have been experimenting with 5-7cM segments. Probably 95% do not triangulate on either side, and therefor must be IBS. It appears the remaining 5% may be IBD - they "fit" into existing Triangulated Groups, and have not been found to be wrong. All triangulations should be formed (based, grounded) on three widely separated kits. Close relatives have identical SNPs over long regions - it's like triangulating with an identical twin with the same DNA. Distant cousins increase the probability that you will weed out IBS segments. With known cousins you are starting with only one string of known SNPs among them over the TG. So be I am very suspect of triangulation below 5cM. We really need some good studies with phased data... Jim - www.segmentology.org > On Dec 3, 2015, at 10:01 PM, "ahnen@awest.de" <ahnen@awest.de> wrote: > > Yes but it still means that if there was a 8cM ancestral segment and it gets cut into half the size (just an example, recombination is basically randomly happening) for the next generation it's still a small portion of our ancestors DNA. > > However for all testing companies and everyone using the minimum 5, 7 or even 10 cM rule it doesn't exist anymore. > > But it does exist still. Given the enormous number of IBS/IBC at that range it doesn't make sense to follow them as a general rule but in case when a relationship is know from paper trail it makes sense to extend the search among a group of known cousins further. > > Andreas > >> On Dec 4, 2015, at 05:24, Jim Bartlett <jim4bartletts@verizon.net> wrote: >> >> Lindsey >> >> Our paper trail means nothing to the DNA. The DNA passes down just the same for rich/poor, famous/not-so, known relatives/not known. Just because we know the cousin, doesn't, somehow, change the quality of segments. The average ratio of IBD/IBS at any cM value will not change. >> >> Jim - www.segmentology.org >> >>> On Dec 3, 2015, at 3:06 PM, Lindsey Britton <lplantagenet@aol.com> wrote: >>> >>> I will definitely concentrate on the large segments. I was just wondering whether it was reasonable to assume small segments were likely to be ancestral, too, once the relationship and common ancestors had been identified. >>> >>> Lindsey >>> >>>

Well Jim, I have a case with 3 different individuals (plus one additional sibling and myself as a son) where we all match through the same paper trail to a common ancestor and we also have a triangulated segment that for some of us is below 5 cM. There are also several other matches that triangulate as well but I haven't contacted them yet to identify their path to the common ancestor. It's tedious work that took me almost a full day but it's possible. Andreas > On Dec 4, 2015, at 22:10, Jim Bartlett <jim4bartletts@verizon.net> wrote: > > It is true that lots of 4cM IBD segments exist for all of us. But there are two problems: > 1. However many there are, you would be swamped with many more IBS segments. > 2. We don't have a good way to test for IBS, short of phasing which most folks can't do. > Refer to my blog. All the segments in your body (on chromosomes in your DNA) are IBD! I call them ancestral segments. You only get your DNA from your ancestors - 1cM segments, 3cM segments, all of them. The segments we "see" in a chromosome browser or a spreadsheet are shared segments (with a Match), as determined by a computer program comparing your SNPs (from both your parents) with your Match's SNPs (from both of his/her parents). This shared segment is made up, fabricated. Over the short term - with only a few hundred SNPs - it's easy for the computer algorithm to "stitch together" apparently matching segments. > As much as I tout triangulation, we don't really know if it works for very short segments. I'm comfortable predicting that triangulation works to sort out IBS segments over 7cM virtually all the time. These are typically longer segments, and I have yet to find any which don't work over 7cM. I have been experimenting with 5-7cM segments. Probably 95% do not triangulate on either side, and therefor must be IBS. It appears the remaining 5% may be IBD - they "fit" into existing Triangulated Groups, and have not been found to be wrong. All triangulations should be formed (based, grounded) on three widely separated kits. Close relatives have identical SNPs over long regions - it's like triangulating with an identical twin with the same DNA. Distant cousins increase the probability that you will weed out IBS segments. With known cousins you are starting with only one string of known SNPs among them over the TG. > > So be I am very suspect of triangulation below 5cM. We really need some good studies with phased data... > > Jim - www.segmentology.org > >> On Dec 3, 2015, at 10:01 PM, "ahnen@awest.de" <ahnen@awest.de> wrote: >> >> Yes but it still means that if there was a 8cM ancestral segment and it gets cut into half the size (just an example, recombination is basically randomly happening) for the next generation it's still a small portion of our ancestors DNA. >> >> However for all testing companies and everyone using the minimum 5, 7 or even 10 cM rule it doesn't exist anymore. >> >> But it does exist still. Given the enormous number of IBS/IBC at that range it doesn't make sense to follow them as a general rule but in case when a relationship is know from paper trail it makes sense to extend the search among a group of known cousins further. >> >> Andreas >> >>> On Dec 4, 2015, at 05:24, Jim Bartlett <jim4bartletts@verizon.net> wrote: >>> >>> Lindsey >>> >>> Our paper trail means nothing to the DNA. The DNA passes down just the same for rich/poor, famous/not-so, known relatives/not known. Just because we know the cousin, doesn't, somehow, change the quality of segments. The average ratio of IBD/IBS at any cM value will not change. >>> >>> Jim - www.segmentology.org >>> >>>> On Dec 3, 2015, at 3:06 PM, Lindsey Britton <lplantagenet@aol.com> wrote: >>>> >>>> I will definitely concentrate on the large segments. I was just wondering whether it was reasonable to assume small segments were likely to be ancestral, too, once the relationship and common ancestors had been identified. >>>> >>>> Lindsey >>>> >>>>