A brief note that might be of interest.
I attended an excellent seminar yesterday on the evolution, development and function of neural crest cells.
One of the issues that came up is what we mean by "conservation of function" in evolution.
In brief: a gene may conserve its function by some measures and not others. In evolutionary time, the same function cana be mediated by different genes, and the same genes may alter their function. Its in part because genes have neighbors. Its in part because function itself is open to many measurements. The very distinction between analogy and homology in evolutionary terms can become moot as we move from one measurement frame to another. Its not as simple as the textbooks imply!
Here is a simple analogy. To what extent does an egg have the "same function" if I use it to cook different dishes? Think about it. What do I MEAN by function here? Of course cultural evolution has now made egg substitutes as well.
One egg can do many things (cf. gene). One function can be served by either an egg or an egg substitute (cf. genes).
The mouse gene in a human may, in some senses, have the same function as the gene does in the mouse. But not necessarily, at least by some measures.
I ran into this problem recently with some work we are doing with mice that have the human Huntington's gene. In some ways this transgenic animal seems to be a good model for Huntington's (e.g. striatal damage), but in other ways it does not. The gene is working on a mouse brain in one case and a human brain in another case. That's a different set of worlds the gene is asked to live in, and contribute to.
And, many of these special genes have major or minor effects (by some measures) if the mice are given different environmental histories (experiences).
The gene to behavior link is both powerful, and very very indirect. Therein lies the challenge for research.