The Molecular Graphics Laboratory Forum

Hi folks,
I'm setting up a docking simulation in which I want to include one side chain (a Leu residue) as flexible. I prepared the .gpf and .dpf files of my system using AutoDockTools version 1.5.2 (compatible with AutoDock 4.0) installed on a PC with Windows XP. To prepare the files, I followed instructions that I found in the tutorial "Using AutoDock 4 with AutoDockTools" written by Ruth Huey and Garrett Morris. When I finished creating the .gpf and .dpf file, as usually, I transferred them to a Linux cluster to make AutoGrid+Autodock calculations (sequentially). Autogrid calculations produced the maps and all other files, but when I launched Autodock, an error returned saying that a map was not present. I checked the two files and found a strange thing. In the .gpf and .dpf files the ligand atom types do not correspond. These are the lines for the .gpf file:
ligand_types A OA Cl # ligand atom types

and these are the lines for the .dpf file:
ligand_types A OA Cl C # atoms types in ligand
Practically, it seems to me that in this file the flexible atom of Leu was merged to the ones of the ligand. As a consequence, in the .dpf file there is the need for the
map hAHRrigid.C.map # atom-specific affinity map
that has not been previously calculated by AutoGrid, since the C atom was not (correctly) included in the ligand atom type.
I tried to edit the .dpf and .gpf files in two ways, in one case erasing in the .dpf the line containing the map hAHRrigid.C.map and the C in the ligand atoms types, in another case adding in the .gpf file the C to the ligand atom types, in order to have the map hAHRrigid.C.map. I expected that in the first case the Leu residue would have been rigid, and in second case flexible; instead, in both cases it was flexible.

I really don't understand the reasons for this strange "behaviour". Can anyone help me?
Thanks in advance

Hi,
can anybody help me with the above question?

AutoDock requires a grid map for each type of atom whose movements are
searched during the docking calculation. The moving atoms are all the atoms
in the ligand PLUS those in the flexible residues IF the experiment includes
flexible residues.

Your error message indicates that there is one or more atoms of type 'C' in
the flexible residues but none of type 'C' in the ligand.
When you are modeling flexible residues, the atom types in the flexible residues
must be added to those in the ligand in creating the gpf so that AutoGrid
will create all the map files for necessary for the AutoDock calculation.

To do this in ADT, use Grid->Set Map Types two times before writing the gpf file:
-FIRST click either Choose Ligand... or Open Ligand....
<choose your formatted ligand pdbqt filename>
-THEN either Choose FlexRes.... or Open FlexRes
<choose your formatted flexres pdbqt filename>

When you have written the gpf, look at it to make sure that C is in the list
following the keyword 'ligand_types' AND that the gpf contains the line:
map hAHRrigid.C.map
<to be sure that the atom types in the flexres have been added>

SIMILARLY, in the Docking menu in ADT
Docking->Macromolecule->Set Rigid Filename...
< here pick hAHRrigid.pdbqt>
Docking->Macromolecule->Set Flexible Residues Filename...
< choose your flexres file ??hAHflex.pdbqt??>

Best wishes,
Ruth