A comprehensive, non-redundant composite protein sequence database is
described. The database, OWL, is an amalgam of data from six publicly-
available primary sources, and is generated using strict redundancy cr
iteria. The database is updated monthly and its size has increased alm
ost eight-fold in the last six years: the current version contains > 7
6000 entries. For added flexibility, OWL is distributed with a tailor-
made query language, together with a number of programs for database e
xploration, information retrieval and sequence analysis, which togethe
r form an integrated database and software resource for protein sequen
ces.