Kuntz Home / DOCK Home / DOCKumentation Contents / DOCK

distmap

Brian Shoichet, Dale Bodian, Elaine Meng

Overview / Input

Overview

distmap creates the grid file necessary for contact scoring. Contact scoring using a grid is much faster than the contact scoring implemented in earlier versions of DOCK, in which pairwise ligand-atom/receptor-atom distances are evaluated for all ligand orientations. distmap produces a file of precalculated "scores" for points in a cubic grid. Each point is evaluated in terms of its distance to each of the receptor atoms included in the calculation. Negative scores occur when a point is too close to a receptor atom, as defined by the limits ccon and polcon (for receptor nonpolar and polar atoms, respectively). If there are no "bad contacts," the score for a given point is the total number of receptor atoms within discut Ångstroms of that point. Only non-hydrogen atoms are considered. During docking, each ligand atom receives the score of the nearest point, and the total contact score for the orientation is the sum of the scores of the ligand atoms. We caution that the grid-based contact scores and the original contact scores are not identical. The original scoring scheme is available (see oldscore).

Input

The input file names and parameters are read from a file called INDIST, which should not contain any blank lines:

variable name	Fortran format	example
pdbnam	A80	`2ptc.pdb`
scoren	A80	`2ptc.map`
polcon	F	`2.3`
ccon	F	`2.8`
discut	F	`4.5`
perang	I	`3`
[boxlin]	A80	`box.pdb`

pdbnam: is the name of the PDB-format file containing the receptor atoms which will be used for calculating the contact scores. All or a portion of the receptor may be included. Currently, any hydrogen or deuterium atoms are ignored. The advantage in using only a portion of the receptor is that greater resolution may be attained without increasing the number of grid points. The grid is automatically constructed to enclose all of the nonhydrogen receptor atoms in the file, unless a limiting box is specified (see boxlin below).
scoren: is the name of the output grid file.
ccon: is the close contact limit, in Ångstroms, for receptor nonpolar atoms. Grid points within ccon Ångstroms of any receptor atom that is not a nitrogen, oxygen, hydrogen, or deuterium are given a negative score.
polcon: is the close contact limit, in Ångstroms, for receptor polar atoms. Grid points within polcon Ångstroms of any receptor nitrogen or oxygen are given a negative score.
discut: is the cutoff distance, in Ångstroms, for "good" contacts. As long as a grid point does not violate any close contact limits, its score equals the total number of receptor atoms within a radius of discut Ångstroms.
perang: is the number of grid points per Ångstrom. Values from 1 to 10 are accepted, as long as the total number of grid points is not too large. The total number is controlled by the spacing of the points and the volume enclosed by the grid as a whole. The volume is controlled by the atoms in pdbnam unless a limiting box is used. If the total number of grid points exceeds array bounds, one can decrease the volume of the grid, decrease the points per Ångstrom, or recompile with larger array dimensions. Caution is warranted in increasing array dimensions, since exceeding physical memory will result in page faulting and slow DOCK runs.
boxlin: is the name of the PDB-format limiting box, created using showbox. This is optional. If boxlin is specified, only receptor atoms enclosed in the box will be used in the calculation. This option allows the volume of the grid to be less than that needed to completely enclose the atoms in the file pdbnam.

Curator: Daniel Gschwend, gschwend@cgl.ucsf.edu (rev. 1 September 1995)