As the natural continuation of the discussion about the X-axis in chromosome painting graphics that appeared on one of these DNA lists recently, I would like to ask the obvious question: How are the (combination of reference population probabilities) values assigned to the Y-axis calculated? The reason I ask is because I have triangulated a (10+ cM) segment of my dad's chromosome 2 to his paternal line (viz., it descends from a so-many-greats Bonner ancestor). At about the 160M mark in the Dodecad K12b calculator (e.g.), there is a big peak of Southwest Asian - so much so that it crowds out all the other populations to exclusivity for about 1M bp (gedmatch F209176, if you want to follow along). Now I get a list of all the SNPs in that region and check out the SNP reference similar to this: http://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?rs=[PUT rsID HERE]#maintab And go down to "population diversity" and check out the numbers... I was surprised that there is very little difference between the populations, and I don't see how one could get a near 100% (as depicted in the chromosome painter) from the values in the population diversity sections, or anything like it. Are the populations shown in the chromosome painter dependent on adjacent SNP population prediction, or are the populations assigned to each SNP independent? Is this the product of serious, high-level statistical "teasing" of the data? Is there any place online that shows a sample calculation for one or more SNPs? I'd like to follow along before I make comments about how my dad's DNA is partly Asian in origin. When I look at the population diversity information, those SNPs look just as European to my untrained eye. Best, Gregg Bonner