informative sites & bootstrap

Brian Foley btf at t10.lanl.gov
Thu May 7 14:17:47 EST 1998


Warren Gallin wrote:
> ...this discussion is only about MP analysis.
> Although I see the point of your discussion above, it is 
> not relevant to a MP approach; it is appropriate to a ML 
> approach and distance approaches.

	I agree.  I was thinking the thread was more general
about bootstrapping.  Bootstrapping can be used to generate
data sets for ML and other distance approaches as well as
for MP.
	In my experience, the main point is to find the
tree which truly reflects evolutuion.  And given that we often
do not know the true evolutionary history, we instead get 
a tree and want to know how "robust" this tree is.  Bootstapping
is one method of testing the strength of the data. 
	Another method of determining how confident we
are in a given tree topology, is to use different methods
such as MP, ML and neighbor-joining.  If we have enough good
data, and we use all methods correctly, then each should
give a very similar tree topology.  When we start getting 
different topologies with different methods then we have
to question which one is the best tree, and what types of
data are likely to work well with each method.
	While it may be incorrect to speak of "informative
and non-informative sites" in a distance-based phylogenetic
method, the concept, if not the technical term, still applies.
We can have sites that contribute to similarity (distance)
via convergent evolution rather than lack of divergence.

 ____________________________________________________________________
|Brian T. Foley               btf at t10.lanl.gov                       |
|HIV Database                 (505) 665-1970                         |
|Los Alamos National Lab      http://hiv-web.lanl.gov/index.html     |
|Los Alamos, NM 87544  U.S.A. http://www.t10.lanl.gov/~btf/home.html |
|____________________________________________________________________|




More information about the Mol-evol mailing list