Article ID: 598 - Last Modified: December 4, 2010
Is there a way to generate grids for a large number of pdbs? Also, is there a way to extract the ligand easily, so that it does not become part of the grid?
On the Script Center there is a cross-docking script (xglide.py) that can be used to prepare receptors/complexes (with the command-line 'prepwizard'), and extract the ligand and receptor structures prior to running Glide Grid Generation and Ligand Docking.
It is rather difficult to reliably extract the ligands automatically before generating grids. The XGlide program by default assumes that the first ligand-size molecule in a complex is a ligand and extracts it. You can't supply an alternative ASL expression for identifying the ligand, but you can specify an explicit molecule number for the ligand (on the COMPLEX line of the input file).
As an alternative to letting XGlide split the complexes, you could use the command line pv_convert.py script, also from the Script Center. This script can split a complex into ligand and receptor structures (the -l and -r options), using an ASL expression to identify the ligand. By default, the script assumes that the last molecule in the structure is the ligand, but you can instead specify an ASL expression for the ligand (the '-a' option). One generic sort of ASL expression for detecting a ligand might be '(mol.atoms 5-130) AND NOT (( backbone or atom.pt 2HA, 3HA ) OR ( sidechain ))' for ligand-sized molecules that aren't peptides.
Keywords: batch grid generation, pv_convert.py, xglide.py, script, extract ligand, process large numbers of structures.
Type the words or phrases on which you would like to search, or click here to view a list of all
Knowledge Base articles