Monday, January 24, 2011
Joe Felsenstein just gave a really nice talk at phyloseminar. If you missed it live, the talk was recorded and will be archived.
I liked the talk because it gave us some hints about the future of comparative methods. Often it takes the field of comparative biology 20 years to catch up with Joe, but perhaps this time we can shorten the interval a little bit.
The talk included discussion of inferring the evolution of geometric shape on trees, placing fossils on phylogenies using likelihood, and applying threshold models to comparative data. Everything was placed in an historical context, which was nice. I particularly appreciated funny snippets about an argument between Felsenstein and Bookstein, Felsenstein's take on the famous Wright "guinea-pig-as-blackboard-eraser" story, and a really interesting idea about QTLs. The above image is from the movie G-Force - it's terrible, please promise not to watch it.
Felsenstein concluded with two crucial points. First, we're witnessing what he called a "grand reunion" of quantitative genetics and statistical comparative methods - fields that have remained too separate for too long. I will make a similar point in my talk. Second, you need more tips! New comparative methods are data-hungry and even 100 taxon trees can be barely big enough for some methods.
My phyloseminar is up next on Feb. 24 - please tune in! Also I've started a twitter account @lukejharmon.