3D Protein Surface Segmentation through Mathematical Morphology

Abstract

An important research activity in proteomics is the study of geometrical and topological aspects that characterize the surfaces of proteins or ligand molecules. The functionalities and the molecular interactions of an unknown protein can be, in a first approximation, inferred by comparing her geometrical shape with that of known molecules. The aim of this paper is to identify particular features of molecular 3D surfaces in a new and efficient way by using a set of mathematical morphological operators. The obtained segments constitute pockets and protrusions. Pockets are generally the concave and deepest spots were active sites take place. The protrusions set, on the other hand, is the complementary region of the pockets and therefore, can be used by docking algorithms in the study of protein-protein and protein-ligand interactions.