Genes, a gene, genes....

I spend some time thinking about genetic architecture. If you read media accounts of genetics you wouldn't know there was such a thing. Why not? Because it makes a narrative rather messy and complex. Rather, a better story is one where you have a gene-for-something. But the gene-for-something model is probably most useful in the domain of recessive diseases. In these cases you have a mutant variant which has a powerful effect when found in two copies in an individual. The distinction between the disease state and the "wild" state is rather straightforward. Not so with many "complex traits." Two papers came into my RSS feed which illustrate two ends of this dynamic. First, A Sexual Ornament in Chickens Is Affected by Pleiotropic Alleles at HAO1 and BMP2, Selected during Domestication: