Pocket extraction on proteins via the Voronoi diagram of spheres

J Mol Graph Model. 2008 Apr;26(7):1104-12. doi: 10.1016/j.jmgm.2007.10.002. Epub 2007 Oct 7.


Proteins consist of atoms. Given a protein, the automatic recognition of depressed regions, called pockets, on the surface of proteins is important for protein-ligand docking and facilitates fast development of new drugs. Recently, computational approaches have emerged for recognizing pockets from the geometrical point of view. Presented in this paper is a geometric method for the pocket recognition which is based on the Voronoi diagram for atoms. Given a Voronoi diagram, the proposed algorithm transforms the atomic structure to meshes which contain the information of the proximity among atoms, and then recognizes depressions on the surface of a protein using the meshes.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Algorithms
  • Binding Sites
  • Computer-Aided Design
  • Drug Design
  • Ligands
  • Models, Molecular*
  • Molecular Structure
  • Pharmaceutical Preparations / chemistry*
  • Pharmaceutical Preparations / metabolism
  • Protein Binding
  • Protein Conformation
  • Proteins / chemistry*
  • Proteins / metabolism
  • Surface Properties
  • Terminology as Topic


  • Ligands
  • Pharmaceutical Preparations
  • Proteins