One of the problems now is if you use SNP statistics for Germans you have to take into account the N% of Turks in that population which changes over time, with DNA they really should be separate groups. The assumption that the genes will blend doesn't really pan out quickly enough for it to be ignored in the near term.
Race was a crude approximation and better techniques are available today. There is a push to use a wider variety of reference genomes which makes sense especially now that that computers are more powerful. There seems to be an assumption by others here that going from an unsophisticated crude approximation to a sophisticated one will somehow validate their other assumption that 'race is only skin deep'. I am not as optimistic.
Race was a crude approximation and better techniques are available today. There is a push to use a wider variety of reference genomes which makes sense especially now that that computers are more powerful. There seems to be an assumption by others here that going from an unsophisticated crude approximation to a sophisticated one will somehow validate their other assumption that 'race is only skin deep'. I am not as optimistic.