Abstract
In 1991, we introduced a family classification system for glycoside hydrolases and transglycosidases based on amino acid sequence similarities. More recently we have introduced a similar classification system for glycosyltransferases. With the growing amount of sequence and biochemical data, the number of families of glycoside hydrolases and glycosyltransferases has grown over the years. There are presently over 80 families of glycoside hydrolases and more than 50 families of glycosyltransferases. These families can be accessed conveniently on a permanently updated www server at URL: http://afmb.cnrs-mrs.fr/similar topedro/CAZY/db.html. Although it is likely that other families will be discovered, it is reasonable to assume that the most important families are already identified and that general inferences on the role and importance of carbohydrates in different organisms can be made. An advantage of the sequenced-based families is that they can be readily searched in genomes, allowing a global and detailed comparison of the glycoside hydrolase/glycosyltransferase repertoire of various organisms at a genomic scale. Here we have searched glycoside hydrolases and glycosyltransferases in 18 completely sequenced genomes covering Archaea, Bacteria and Eukaryotes.