Mailbag: Pearls for the swine

Re: “Genetic lapidaries”:

Hope all is well in Wisconsin. Regarding your post on the new Southern African genomes and the tendency to discuss phenotypic associations... Isn't this just a product of the priveleged status genetic technology gets in general and the way in which students are genetics in introductory biology courses? Students get taught that genotype > phenotype and get great Mendelian examples which make the process very comprehensible even if they actually underplay to the point of misinforming the students the actual complexity of that relationship. Likewise the reporting and general discussion of genetics has always highlighted its knowledge generating capacity - "decoding the blueprint of life"... So when stories like this come out there is a latent expectation to hear about those simple genotype-phenotype associations like the ones you outline. This then becomes self-reinforcing. Obviously these genomes represent a huge amount of information - it is harder to accept and understand that they don't represent the amount of knowledge (at least yet) that we would like them to be.

What you say is true enough, that’s probably where these authors are coming from. I mean, it’s hard to blame the press when the paper serves up these little nuggets of “wisdom”.

Discovering Mendelian associations that differ in frequencies in different populations is a good start. I only object because the ones they’ve “discovered” are the ones that we already knew about!

Anyway, seems to me that the thing to do is figure out how many of the “damaging” amino acid substitutions actually are new things in the (European) reference sequence that have been selected recently. We may have a bit of a statistical problem there – if you want to test hypotheses about 13,000 or so amino acid variants, the Bushmen are not a very large sample…