tag:blogger.com,1999:blog-7785493.post4642956726767365604..comments2024-01-04T04:11:55.717+02:00Comments on Dienekes’ Anthropology Blog: Estimating TMRCA for a pair of Y-STR haplotypes using Average Squared Distance (ASD) cont'dDienekeshttp://www.blogger.com/profile/02082684850093948970noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-7785493.post-80418881766063389662008-10-13T18:04:00.000+03:002008-10-13T18:04:00.000+03:00I applaud your modelling efforts and have said so ...I applaud your modelling efforts and have said so before. I also respect Jim Heald. My comment is how real is this simulation?? 1. P(g): I use the number 30 for years to gen conversion. At this point each 1 year "error" is about 3%+ error in estimating time to a specific occurrance. Since events are usually dated by year, the spread in years can be very large. 2. You use a fixed number for mutation rate; I estimate as much as a 30:1 range over the "slow" mutators and approaching 100:l over all dys loci. 3. I have studied the Ian Cam of the FtDNA Clan Gregor extensively. They all have a common ancestor who was born about 1300 AD. (note there have been 24 chieftains since him or an average number of years per chieftain of c. 30). The range of the number of mutations of this data set ranges from: 0 for the clan chieftain, a direct descendant to 7 for some some cousins whose ancestors travelled to Australia. (this is over 37 dys loci). How would one ever estimate the number of g's back to a common ancestor??<BR/><BR/>I have come to the conclusion that, "weak" as it is, TMRCA is the best approach and the larger the data set the better the estimate. Somehow?? on an individual basis, this doesn't make much sense due to multiple step mutations and apparent increase in rate with change of country?? (climate?). The only way it seems to work is if you use the group of descendants together?<BR/><BR/>So I can demonstrate with just one data set that "modelling" the mutation process still requires some inputs we don't understand or have descriptions of??<BR/><BR/>Maybe you can argue that my examples are all acceptable results within the framework of your model?? I just don't see it.McGhttps://www.blogger.com/profile/03459589185170647441noreply@blogger.com