« The Biological Basis of Multiple Orgasms | Gene Expression Front Page | Carl on language »
February 25, 2005

Gene-environment interaction and gene-environment correlation

In my previous post on gene-environment interaction David pointed out that I unfortunately confounded the term with gene-enviornment correlation. Vge should have been decomposed into Vge and Vcov(g,e) (covariation between environment and genes).

Gene-enviornment interaction may be conceived of as the influence that different environments may have on different genotypes in shaping the developmental arc of the phenotype, for example, norm of reaction. In this case, genotype A and B might result in the same phenotypic consequences (or at least the same variation of consequences) in environment X, yet still diverge greatly in their expression in environment Y. One common example often given is the growth of two bacterial strains in two mediums, in one medium both may show great vigor, but in the other medium one medium may be far superior.

But for most of the earlier post I was explicitly talking about gene-environment correlations which increase the variance of the phenotype. It seems to me that this process is easier to generalize about than gene-environment interaction, which would have varied contextual manifestations. It is the sort of thing that social engineering might find tractable, as opposed to gene-environment interaction which might elude one-size-fits-all policy prescriptions.

If you are curious, I suggest chaper 18 (PDF) of George Carey's Human Genetics for the Social Sciences for more information on this topic.

Posted by razib at 02:32 PM