Glycosyltransferases, the enzymes that build oligosaccharides and glycoconjugates, have received much interest in recent years owing to their biological functions and their potential uses in biotechnology. The analysis of the wealth of sequences that are now available in databases allowed the classification in different families characterized by conserved peptide motifs. Nevertheless, only a limited number of crystal structures is available and molecular modeling appears to be an inescapable tool for rationalization of binding data, engineering of enzyme properties, and design of inhibitors that would be of interest as therapeutic compounds. Because of sequence diversity and limited experimental data, molecular modeling of these enzymes is not straight-forward and utilizes the most recent tools, such as fold recognition programs.
