GENEALOGY-DNA-L ArchivesArchiver > GENEALOGY-DNA > 2011-11 > 1321023272
Subject: Re: [DNA] Correct TMRCA analysis
Date: Fri, 11 Nov 2011 14:54:32 +0000 (UTC)
Didier Vernade wrote:
>Is the Mc Donald tree known by paper records or was it derived by some comparative method ? Marko Heinila is finding differences according to which methods is used to produce the tree.
>Second question : what about Busby et al. ? You didn't react on the many discussions on that paper Latest post being by Mike W with his variance analysis (a few days ago).
The Mc Donald tree was composed by me a couple of days ago from the latest (updated) Donald Clan haplotypes edition, currently 136 of 67 marker R1a1 haplotypes. Many programs exist; however, they do the same thing - they sort out haplotypes (or whatever) based on minimization of the difference between the objects. I do not employ those programs for calculations ( they might be different in that regard), I just use their principal function, which in this particular case is to split the dataset t o a series of branches. The program which I use does it nicely. The "difference" you are talking about might be rotating the tree, for example, and the left side branch next time (when you change setting parameters or add haplotypes) might come up on the right side of the tree, which I do not care about. There are haplotypes which mutations place them in an unstable position between the branches, and they can switch between branches when you add of subtract haplotypes, however, this does not change the pattern and is reflected in the margin of error of the final calculation.
In other words, the tree can be technically different (see above) but branches are practically the same. That is what counts.
"What about Busby et al". What about them? Their paper is very "heterogeneous", and the part with "mutation rates" is a total disaster. In that regard the paper was doomed. Another part about "no geographical trends in diversity" of R1b1b2-M269" in Europe was doomed as well, since they have used bikini-haplotypes. They were not supposed to see anything meaningful. That was the main reason that I "did not react", since I prefer to make positive remarks in discussions, not negative, unless I am challenged. However, in that particular case I WAS challenged, and, as a result, the detailed analysis of the Busby at al. paper is here:
The content of the journal is mainly in Russian, however, the Busby analysis and the following discussion is in English (p. 1831 and further). Technically, the reference is Proceedings of the Russian Academy of DNA Genealogy, vol. 4, No. 9, pp. 1831-1892 (2011).