clustal protein alignment with >50 sequences?

Ashok Aiyar aiyar at ebv.oncology.wisc.edu
Sat Sep 13 11:16:52 EST 1997


On Sat, 13 Sep 1997 00:46:11 GMT,
davesmith at bioch.tamu.edu (davesmith at bioch.tamu.edu) wrote:
>We have 51 different phage and chromosomal sequences of the protein we
>study.  None are more than 200 residues in length.  After an
>exhaustive search of the web, we've been unable to find a program that
>will line up more than 25 proteins at once.  Does anyone know of a way
>to line these up and generate a dendrogram or "family tree" with such
>a large number of individual proteins?

Get the latest version of ClustalW (1.7) or ClustalX (1.5).  They
are available from: ftp://ftp-igbmc.u-strasbg.fr/pub/.  Binaries for
a number of platforms including DOS and MacOS are available from there.

The static arrays used to store alignments (in progress) have been
replaced by dynamically allocated memory.  As a result, there are no
limits on the number or length of sequences that can be aligned - as
long as the computer doesn't run out of memory.  In practical terms,
on a Linux PC with 32 megs of real RAM, and 64 megs of swap space,
I recently aligned 39 sequences that were each approximately 8 kilobases
in length.  The alignment ran for about 3 days, but without any memory 
errors.

I believe that a release of DIALIGN which also uses dynamically
allocated arrays will soon be released by Burkhard Morgenstern and
his collaborators.  DIALIGN is available from: ftp://ariane.gsf.de

Later,
Ashok
-- 
Ashok Aiyar, Ph.D.
McArdle Laboratory for Cancer Research
aiyar at ebv.oncology.wisc.edu




More information about the Bio-soft mailing list