Each data point is a 25-dimensional representation (Global25), and I use Euclidean distance.
There is already https://genetic-distance.com/, where you can look at the closest populations as a table.
The vis should come with many caveats. One is that it is based on the data (and their quality) that is there, which does not cover all populations. Second, the localization of groups is done via LLMs (Claude), and one of the limitations is that a whole population (sometimes as dispersed as "Ashkenazi Jews") is put in one location, sometimes arbitrarily. Third, the interpolation of colours is purely decorative.
Source: https://github.com/stared/genetic-distance-map
With all that in mind, enjoy!
See which populations are most related to Vikings, Mycenaean Greeks, or Canaanites. How about the modern diasporas of Ashkenazim or Roma?