In article <9303150631.AA02073 at net.bio.net> debry at dirac.scri.fsu.edu
Ron DeBry writes:
>The problem with using only 300bp of viral sequence isn't that it is
>only a small fraction of the genome, but that it is a small number of
>bases on an absolute scale.
>>Not all sequences are created equal, for example sequence of the entire
>mitochondrial genome probably wouldn't shed much light on the
>relationships among metazoan phyla, but other factors being equal more
>sequence information is much better than less. Also, it would be best
>to have sequence from different regions of the genome. This would help
>identify cases where the "gene tree" is not the same as the species tree
>(although in practice this is rarely done).
I would like to expand upon this by emphasizing that all genes/gene segments
are clearly not equal in evolutionary terms. Consider highly conserved genes
necessary for viability (eg. low copy histones genes transcriptional regulators
enzymes etc.). Changes within these genes can be highly restricted due to
functional requirements of their encoded proteins.
When mutations within a "conserved" region become fixed within a population,
they can become essentially fixed ("captured") within the genome of that
particular lineage. This is particularly true for, but not necessarily restricted to
insertion/deletions for which the probability of reversion can be extremely low.
I thus believe that even very small gene/protein segments could be used
effectively to produce phylogenetic trees if selected carefully.
brutledge at pnfi.forestry.ca