Thanks for your offer, Tim; It will be good to understand the AncestryDNA process. Like Shannon, I have found little use for AncestryDNA matches - I can't seem to get data for the few interesting Matches, and far too many GEDcom Matches with Ancestry IDs are clearly IBS. It looks like perhaps the flat areas may be generating many Matches at AncestryDNA.... Jim - Sent from my iPhone - FaceTime! On Dec 4, 2013, at 1:09 AM, "Tim Janzen" <tjanzen@comcast.net> wrote: > Dear Jim and Shannon, > There is actually a fairly strong correlation between the cMs and > the number of base pairs throughout the autosomal chromosomes. Last year I > downloaded a series of graphs from Rutgers University that shows this > correlation chromosome by chromosome. I can't seem to find the URL where I > downloaded the graphs from now, but in any case you can download them from > my Dropbox account at > https://dl.dropboxusercontent.com/u/21841126/map%20comparing%20chromosome%20 > position%20to%20cM%20from%20Rutgers.pdf if you want to review them. The URL > for the Rutgers University computation genetics web site is at > http://compgen.rutgers.edu/maps. It describes some of the background there. > In any case, if you review these maps closely you will see that only about 9 > of the chromosomes that have portions of the graph where the slope of the > curve is very flat (in other words there are a large number of base pairs > per cM in these regions of the chromosome). Chromosomes which appear to > have regions where there could be more than 5 million base pairs in a > segment of the chromosome where there is less than 1 cM include chromosomes > 1, 3, 7, 9, 11, 16, 17, 19, and 20. > I suspect that there is something going on here that is more than > just regions where the slopes of these graphs are relatively flat. I > suspect that the phasing of the data by Ancestry.com may be faulty in some > instances. In other words, some of the matches that Ancestry.com is > generating may not be true HIRs that contain 5 million or more base pairs. > I expressed some concerns about the quality of the phasing that Ancestry.com > is doing in an analysis I did this summer which I posted at > http://archiver.rootsweb.ancestry.com/th/read/AUTOSOMAL-DNA/2013-08/13765547 > 77. We would need to review some of the raw data files to sort this out. > Shannon, if you could send me some raw data files that appear as matches in > Ancestry.com but don't meet the 1 cM threshold in GEDmatch, I would be happy > to review the files and determine what the lengths of the longest HIRs are. > Sincerely, > Tim Janzen