At my other weblog I just posted something about the possibility for fitness benefits for those who are heterozygotes for the GJB2 locus. For one particular allele GJB2 causes deafness when homozygous. About 1 out of 35 southern Europeans, 1 out of 80 northern Europeans, and a large number of Ashkenazi Jews are carriers, that is, phenotypically wild type heterozygotes. Non-European populations seem to lack this allele. The south-north gradient seems to hint at long term density dependent effects, but a recent presentation at a European genetics conference suggests that carriers might heal their wounds faster! In other words, carriers might be somewhat more fit than either homozygotes, obviously more fit than the deaf, but perhaps more fit than wild type homozygotes. This is called overdominance, and the classic situation is sickle-cell anemia, where the heterozygote is protected to a great extent against malaria. It doesn't matter what the GJB2 mutation confers, but the pieces are coming together. I think many "disease" like eczema and cystic fibrosis are important because as Greg points out the frequency of the alleles which cause these problems are just too high to be attributed to mutation-selection balance (my understanding is that the rule-of-thumb for mutation rate at a Mendelian locus is on the order of 10^-8, and if the a lethal recessive's equilibrium frequency is the square root of the mutation rate, we're talking 1 out of 10,000, not 1 out of 100 as carriers).

This stuff is important because if Moysiz and Pritchard papers are correct, that there is a lot of recent human selection, a subset of these are going to be overdominant alleles which haven't had time to acquire modifier loci which mitigate their downsides (whether as heterozygotes or as homozygotes). So, Mendelian diseases are important because people get sick, but they might also be a nice sketch outline of the map of recent human selection....