Biology Reference
In-Depth Information
speciation; orthologs are derived from a common ancestor and tend to have
similar functions.
Paralogs
are homologous sequences produced by gene
duplication and represent genes that duplicated within an organism and then
diverged, often having different functions.
Xenologs
result from the horizontal
transfer of a gene between two organisms, with the function being variable,
although often it is similar. Sometimes divergent lineages are morphologically
similar (
homoplasy
). Homoplasy can occur due to a reversal to an ancestral trait
in a lineage or to independent evolution (convergence) or to parallelism (simi-
larity resulting from the same developmental genetic mechanisms) (
Wake et al.
2011
).
Inaccurate trees also may occur due to a phenomenon called
long-branch
attraction
(
Bergsten 2005
). When long branches are in close proximity to short
branches on evolutionary trees, maximum parsimony will recover the wrong
tree because the long branches tend to group together or “attract each other.”
Methods to avoid long-branch attraction include excluding faster evolving third-
codon positions, excluding long-branch taxa, and sampling more taxa to break
up long branches, as well as sampling more characters.
The statistical methods used for developing phylogenetic trees involve scor-
ing for optimality in some way. Maximum parsimony (MP) and maximum likeli-
hood (ML) methods attempt to find the tree or trees that optimize tree length
or likelihood of the DNA sequences evolving along the trees. Bayesian methods
consider many possible trees. A problem with all these trees when using large
data sets, including many gene sequences that have been concatenated or com-
bined or the use of entire genomes, is that many possible trees can be produced.
Other concerns include the problems of discriminating between homology and
convergence or parallel evolution (
Wake et al. 2011
). How to best handle enor-
mous datasets remains to be resolved, although research is being conducted to
do so (
Sanderson et al. 2011, Xia and Yang 2011
).
Another issue that can complicate phylogenetic analysis is missing data
(
Sanderson et al. 2011
). Missing data can make the tree building more complex,
producing “tree terraces.” The solution to such terraces is to focus on filling key
data gaps to improve the effectiveness of tree-building programs.
12.6.9 Software Packages
Numerous software packages for phylogenetic analyses are available and sup-
ported (
Swofford and Olsen 1990, Eernisse 1998
). Software evolves rapidly and
several software packages make it easy to conduct analyses using multiple meth-
ods. If the results are compared and there is concordance among the different
