3D motifs

Abstract

Three-dimensional (3D) motifs are patterns of local structure associated with function, typically based on residues in binding or catalytic sites. Protein structures of unknown function can be annotated by comparing them to known 3D motifs. Many methods have been developed for identifying 3D motifs and for searching structures for their occurrence. Approaches vary in the type and amount of input evidence, how the motifs are described and matched, whether the results include a measure of statistical significance, and how the motifs relate to function. Compared to algorithm development, less progress has been made in providing publicly searchable databases of 3D motifs that are both functionally specific and cover a broad range of functions. A roadblock has been the difficulty of generating detailed structure-function classifications; instead, automated, large-scale studies have relied upon pre-existing classifications of either structure or function. Complementary to 3D motif methods are approaches focused on molecular surface descriptions, global structure (fold) comparisons, predicting interactions with other macromolecules, and identifying physiological substrates by docking databases of small molecules.