A paper was recently published in which it was demonstrated that in
phylogenetic analysis, a potentially misleading resolution could be
obtained in cases where the sampled genes varied greatly in their GC
content. The conclusion was, as I recall, that "high-GC" genes were
spuriously linked by virtue of this compositional bias by itself. I'm
going crazy trying to recall what this reference was..... can anyone
help?
Thanks, Jonathan Wendel
