Hi Edward, I just need to comment on this. To define contact atoms, you must use "solvated united atoms" that is atoms with implicit hydrogens where needed (united radius), this radius being further increased by that of a molecule of water. Why is that? In vacuum, it was shown that at a distance twice the sum of their vdw radii, atoms still attract each other. To take into account the screening effect of water, it has been convened that if a water molecule could fit/pass between two atoms, then the vdw interaction between them would fall off to zero. Therefore, by increasing united radii with the radius of a water molecule, it is easy to calculate the distance, Dij, between atoms i and j, and figure out that when Dij >= Ri + Rj + 2 R(H2O), then there is no vdw interaction, that is no contact. Otherwise, contact exists if Dij < Ri + Rj + 2 R(H2O). Note that in this case, the solvated united atoms intersect and the surface of the intersection disc would be proportional to the strength of interaction between these atoms. In which case the vdw contact refers to the vdw interaction rather then the vdw radii. "Cracks" between atoms correspond to "reentrant" surfaces as defined by Fred Richards.
Best regards, Nadir Pr. Nadir T. Mrabet Structural & Molecular Biochemistry N-gere - INSERM U-964 University of Lorraine, Nancy School of Science and Technology and School of Medicine 9, Avenue de la Foret de Haye 54500 Vandoeuvre-les-Nancy, France Tel. (direct) : +33 (0)3.83.68.32.73 Tel. (secretary) : +33 (0)3.83.68.32.92 Cell. : +33 (0)6 11 35 69 09 Fax. : +33 (0)3.83.68.32.79 Email : Nadir.Mrabet <at> univ-lorraine.fr ----- Le 15 Mai 15, à 21:07, Edward A. Berry <ber...@upstate.edu> a écrit : > If I remember correctly, there are two different ways to calculate a surface > by > rolling a ball over it, and i think that I want a program to calculate the > non-conventional one. > As I understand, the ASA is defined as the surface traced out by the _center_ > of > the rolling sphere, i.e. one radius above the vdw surface. The justification > being that an atom (i.e. it's center coordinates) can't get any closer to the > surface than that. The second type of surface is defined by the closest > approach of the _surface_ of the rolling sphere, i.e. it would be the vdw > radius of the (protein) but not descending into cracks between atoms where the > rolling ball won't fit. > For making models of a multisubunit protein, and wanting to be able to > assemble > the separately-made subunits so that they make intermolecular vdw contacts, > the > second kind of surface would be desirable, as otherwise atoms won't be able to > get their surfaces closer than twice the sphere radius. Is that an option that > can be chosen in pymol or such? > Thanks, > eab
