Hi Doug, RE:  > "How small a block should I trust," You should trust down to 1 cM, if you have an interest in delineating a comparison to separate haplotype groups. I would say forget what you have been told about HIR less than 7 cMs, and remember what you know about Y-DNA - that SNPs want to hang around forever. That means HIRs also want to hang around forever. My thoughts are that the segments under 7 cMs tell us most of what we need to know about distant relationships. You should concentrate on segments less than 7 cMs. RE:  > "This is because the "other person" would have some markers that are homozygous and don't match the  same side as the supposedly matching side of me." I think the normal amount of recombination is in the 8% range. Other than recombination, if you are looking at inherited segments that are ancient, then you should expect to see some strange stuff. So, I would also advise that you forget about what you have been told about cMs and segment age. That only works for the closest relationships, more like a side effect of what is actually going on.  - Dave Hamm RE: On 12/3/2018 1:45 PM, genealogy-dna-request@rootsweb.com wrote: > Date: Mon, 3 Dec 2018 17:46:35 +0000 > From: "McDonald, J Douglas" <jdmcdona@illinois.edu> > Subject: [DNA] HIR size and physical phasing, probability of reality > To: "genealogy-dna@rootsweb.com" <genealogy-dna@rootsweb.com> > Message-ID: <DM6PR11MB3097A06605A0160474A74333BFAE0@DM6PR11MB3097.namp > rd11.prod.outlook.com> > Content-Type: text/plain; charset="us-ascii" > > I have a question. Its usually said that a HIR smaller than say 7 cM is frequently not really from > shared descent. > > I presume this means that just by accident its half identical ... and that if we actually had truely > phased (from a physical measurement, which I have of myself) data of one person we could be > surer of a real match with the other person. This is because the "other person" would have > some markers that are homozygous and don't match the same side as the supposedly matching > side of me. I'm sorry if this is unclear, but that's the best I can say it. > > If indeed this is the case, then smaller segments should be "real". This is important since I > don't have parents available to "genealogically" phase me, and have to use not coly close > cousins but some more distant ones. How small a block should I trust, given that a different LARGE > block on the other person has a known, triangulated relation to me, has the same relation to me? > > > Doug McDonald --