Interpretation of phylogenetic trees

[David R. Nelson]

From: nelsondr at med.unc.edu (David R. Nelson)
Newsgroups: bionet.molbio.evolution
Subject: Interpretation of phylogenetic trees
References: 
Sender: 
Followup-To: 
Distribution: world
Organization: UNC-CH School of Medicine
Keywords: 

I frequently make UPGMA phylogenetic trees of cytochrome P450 protein
sequences and mitochondrial carrier protein sequences.  Among the 
mitochondrial carriers, there is a family of proteins called the 
uncoupling proteins.  These are found exclusively in brown fat 
adipose tissue in mammals in the mitochondrial inner membrane.  In
the tree of these carriers, the closest relatives are 2-oxoglutarate/
malate shuttles, which share 30% sequence identity at the protein level.
A colleague of mine that does not look at phylogenetic trees often,
asked me a simple question that I would like to pose to the network.
He said the tree did not make sense to him because the branch between
the uncoupling proteins and the oxoglutarate malate shuttles was too
deep.  Since the uncoupling proteins only exist in mammals, he thought
the connecting branch should be short, reflecting the recent development
of mammals.  This seemed like a reasonable objection.  The question is
what happened since the original gene duplication, and the development
of the new function of the uncoupling proteins?  Is it probable that the
lineage always maintained some function that was selected for in order
to keep the gene from being lost like present day pseudogenes?  If this 
is correct, how long can a gene be expected to survive in a lineage if 
it does not have a useful function?  A related question may be is the 
transition from one well defined function to another relatively rapid
and how might this be reflected in the branch lengths of a phylogenetic
tree?

Perhaps this will spark some discussion.

David R. Nelson 
UNC Chapel Hill
nelsondr at med.unc.edu