Freitag, 8. Juli 2011

Mei jrine Kusine 1

The main goal of the regband-project was a visual one from the very beginning. How does it look, when multiple matches cluster together, are we able to detect visual pattern, that allow additional interpretation to usual algorythms. Three forms of clusters are quickly distinguishable and seem to be related to different time-frames. Independent from centiMorgan values a cluster, spreading his matches in a wider field, a looser composition ( soft edge ) looks definitely very much younger than those, who have already developed one sharp start- or ending point (hard edge ). Smaller clusters of only three to five MB length with sharp start AND endpoints can perhaps be contributed a considerable age. They are probably real ancestry bands, shared by people of very distant and different countries.
We all have still to learn about the nature of our IBD segments and it would be helpful to clearly distinguish between individual familiar matches, very old ancestry matches, shared by most Europeans and regionally composed stretches, which should have been enriched by specific motives, shaped through natural geographic, ethnic or religious borders up to the complete isolation of a small gene pool.
Ten Ancestry Finder screenshots of Ashkenazim, merged in one.
This is true of the Ashkenazim and knowing, that they show up with about ten times more relatives than a common European, I have to admit – I was much afraid of looking at the first inevitable CSV-files from Ancestry Finder for the regband project. I should have known! At the end of 2010 there were around 300 matches in my own AF-file, the usual range for a continental European. In files of colonial or Finnish descent this number could double, but encountering more than 3000 matches really caused more than one problem. Until then the checking of those usual files took one hour to a whole evening, but now I was busy for more than a week. The threshold problem, the team of 23andme faced in the beginning, showed up as well as a logistic one – namely the difficulty of a clear definition for soft or hard edge clusters. Scrolling up and down an excel-file cannot reveal the true shaping of different clusters, this can only the visual detour of a handmade collection in a chromosome style sheet. Since I don’t work with a computer program, it was impossible and too time consuming to head for all chromosomes, so I simply show the given problems exemplary in the picture of the smaller chromosome 15. On the left side we see a collection of single matches, as explored in two AF-files of full Ashkenazim descent, on the right side the compilation of all the other files.

Keine Kommentare:

Kommentar posten