This is a general annoucement of ProDom release 34 available at:
The Protein Domain database, ProDom, release 34, has been constructed
by clustering homologous segments derived from 53145 non-fragmentary
sequences present in SWISS-PROT 34. It can be retrieved by FTP at:
It provides 18086 mutiple alignments and consensus sequences for
homologous domain families. The vast majority of these families
have been generated automatically using MKDOM, a much improved
version of the DOMAINER program (Sonnhammer & Kahn, 1994, Protein
3:482-492; Gouzy, Eugene, Greene, Kahn & Corpet, in
preparation). Several steps have been taken towards improving the
quality of ProDom. In particular, all multiple alignments have been
recalculated ab initio using the MultAlin program (Corpet, 1988, Nucl.
Acids Res. 16:10881-10890; http://www.toulouse.inra.fr/multalin.html).
In addition a new expertise procedure has been introduced to validate
some domain boundaries.
The Web user interface has also been considerably enhanced.
Links are provided to and from PROSITE and PDB. They have been
with the help of the LASSAP program (http://alize.inria.fr/). Domain
can be searched by keyword, and graphical representations of domain
arrangements can considerably facilitate the structural interpretation
of large protein families.
In addition, we now provide a sensitive homology search procedure
which scans all domain sequences in ProDom and retrieves matches with
only one sequence for each domain family, thus drastically reducing
output redundancy. The most significant matches are visualized
graphically to assist with interpretation. For long queries the former
less sensitive but faster search on consensus sequences is also
provided. Choice is given between the classical NCBI BLAST 1.4.9 and
the new WU-BLAST 2.0a8 allowing for gapped outputs
For any request please mailto:proquest at toulouse.inra.fr
Laboratoire de Biologie Moleculaire
des Relations Plantes-Microorganismes,
Laboratoire de Genetique Cellulaire,
31326 Castanet-Tolosan Cedex