Location of structural domains in proteins

We use surface area measurements based on atomic positions to give a quantitative definition of structural domains in proteins. Segments of the polypeptide chain making a minimum of interactions with the rest of the protein structure are identified on interface area scans, where the area B of the interface between a N-terminal segment of i residues and the complementary C-terminal segment is plotted as a function of i. Domain boundaries appear as minima of B in the scans. The procedure may be iterated to build a hierarchy of subdomains. It detects only continuous domains made of a single stretch of polypeptide chain but may be extended to detect such domains in the presence of discontinuous ones. Domains defined from interface area scans fit very well with globular structural regions identified by inspection of protein models [Wetlaufer, D. B. (1973) Proc. Natl. Acad. Sci. U.S.A. 70, 697-701]. They do not in general correspond to the repeated structural units observed in some proteins by superposition studies. In hemoglobin and hen lysozyme, the domains do not correspond to the coding sequences separated by introns in the genes. © 1981 American Chemical Society.