31 December 2013

More Genetic Ancestry Fun

I reanalysed my 23andMe data using the Dodecad algorithm, this time with the latest Globe13 parameters, and here's what came up:

                            (n=17)    (n=12)
  0.09%  Siberian            0        0
  0.56%  Amerindian          0.4      0.1
  0.00%  West_African        0        0
  0.00%  Palaeo_African      0        0
  2.32%  Southwest_Asian     0        0.5
  0.00%  East_Asian          0        0
 33.12%  Mediterranean       33.7     34.8
  1.00%  Australasian        0        0.3
  0.37%  Arctic              0        0.2
  8.77%  West_Asian          6.2      5.7
 52.50%  North_European      59.1     57.8
  1.27%  South_Asian         0.6      0.7
  0.00%  East_African        0        0

So the column on the far left is my personal output against the notional "population of origin" of bits of my genome (if that makes sense - please attach a pinch of salt to all this; especially the Amerindian bit - can't see that happening, but you never know!). The other columns represent an average of 17 Irish individuals and 12 British (by "ethnicity").

It's difficult to know how much significance to attach to the variations between mine and the average - you can see that if you pool my "Asian" component, you get 12.36, which is a good bit higher than the Irish average (6.8) and the British (6.9), and most of that is at the expense of my "Northern European" component. What are the error bars around this? Is this telling us anything interesting?

In fairness, it probably isn't. Genes are remarkably fluid, and this sort of analysis can really only give us a rough guide. Still, it's fun, and every now and again it may reveal something important... you never know.

No comments:

Post a Comment