Hi,

I have two questions about Branch site REL:

1.  I know this has been asked before, but I want to make sure I understand the answer.  I plan to use Branch site REL for analysis of a number of genes across a group of species.  Am I better of giving my own tree for the analysis based on a well supported species phylogeny (derived from many genes), or is it better to let Datamonkey compute a tree for each specific gene just based on the gene sequences I provide, in which case I might have different trees for different genes?  I do not know for sure that all the gene trees will be the same as the species tree (which is well supported).

2.  Does power to detect selection increase by including more species in the analysis?  Specifically, in this case there are two major clades (Sophophora and Drosophila).  Divergence within each clade is not great, but divergence between the clades is quite large.  Would I be better off just including sequences from the clade containing the species I'm interested in, or is it always better to include as many as I can (assuming I can align the sequences confidently)? 

Thanks so much for your help.

Hi JB,

1). I would use the species tree. Datamonkey builds quick and dirty trees -- they don't make that much of a difference for site-by-site analyses, but could influence branch-specific inference.
2). I would start by just including the sequences of interest. Adding more closely related species is not really going to boost the power to detect selection on the lineages of interest too much.

Sergei

Sergei,

Thank you for the help!