Dear Duncan, I think that it would be particularly interesting to review the raw data files for those people who have matching segments that contain over 2000 SNPs that appear to IBS. I suspect that reason #1 below is rarely the cause for this, except possibly in highly endogamous populations. Reason #2 could be a possible cause for some of these situations. I think that a significant percentage of these situations could be caused by errors in the raw data files that happen to disrupt the run of matching SNPs in the parent but not in the child for the segment in question. You are correct that if we could easily identity IBS segments that meet 23andMe's and Family Finder's match criteria then we could save ourselves the time and effort in trying to identify common ancestors. I have encouraged both 23andMe and FTDNA to phase the data that they can easily phase (using two parent/one child trios). If they did this then matches that are simply IBS could simply be dropped from the match list. Sincerely, Tim Janzen -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of r0berts0n Sent: Wednesday, January 11, 2012 3:53 AM To: [email protected] Subject: Re: [AUTOSOMAL-DNA] SUBJECT: How do you work a 5 way FF matchwith few clues Thanks Tim I'm trying to think of reasons for such long IBS segments and here are my thoughts: 1. There is a very low allele variability in the population in those regions. However if that were so, then lots of people would be seeing those particular segments matching others and I don't think that is so? 2. The segments aren't really that long and Family Finder is tolerating too many mismatches. If that were so the the IBD segments would also be too long and I haven't heard anyone say that is so. I think it is important to understand why this is happening because if we could easily identify IBS segments then we could ignore them and save time and effort in trying to identify matches. Do you have any other ideas for the existence of long IBS segments? Duncan