[Bioperl-l] Re: Comparative genomics

[David J. States]

--On Friday, September 28, 2001 7:00 PM +0100 Elia Stupka <elia@ebi.ac.uk> 
wrote:

>> The key first step is definitely obtaining accurate automated alignments.
>> Without this no amount of tree inference will be sound. It would be a
>> real plus not only to be able to use these for the trees but to be able
>> to see the alignment with the tree object.
>
> That will be taken care of by ensembl-compara the new db that Ewan and I
> are developing. He finished the db, and I now need to finish a way of
> populating it ;)
>
> The compara could hold aligments between sequences as well as alignments
> to HMMs.
>
> Elia

The alignment is as important as the tree finding algorithm, and poorly 
deterimed parts of a multiple alignment can badly skew results.  A number 
of years ago, we wrote a confidence weighted alignment tool to define 
robust parts of the MSA to use in phylogeny and model building.  See

http://stateslab.bioinformatics.med.umich.edu/service/msa/index.html

and

http://stl.bioinformatics.med.umich.edu/software/ctree/ctree1.2/


David



David J. States, M.D., Ph.D.
Professor of Human Genetics
Director of Bioinformatics
University of Michigan School of Medicine
Medical Science Building II, Room 5622A
Ann Arbor, MI   48109
USA
email: dstates@umich.edu
tel: (734) 615-5510
fax: (734) 615-6553
URL: http://stateslab.bioinformatics.med.umich.edu