Recombination, Positive Selection and the Molecular Clock

[Brian Muchmore]

This is a general question, and any opinion is appreciated...

How robust is BEAST in the face of these evolutionary pressures?  It seems to me that allowing for different evolutionary rates among different branches helps to lessen the problem recombination and positive selection pose.  Opinions?  Ideas?

I am trying to model a single gene's evolution using SNP data from 1000 genomes, ancient DNA, etc, but there are mild signatures of recombination and positive selection depending on the population.

[Derek]

Depending on your approach, you might run into some trouble if you compare the 1000 Genomes dataset with another source of sequence data. 1000 Genomes discovers and reports variants randomly and uniformly, which makes it an improvement over HapMap and more amenable to statistical analysis. However it has not discovered all the variants that exist in its samples.