DOCK 3.7 tart: Difference between revisions

From DISI
Jump to navigation Jump to search
No edit summary
No edit summary
Line 26: Line 26:
If you would like to only generate new QNIFFT electrostatic grids, you do not need to run blastermaster. Note, you do need the following input files generated from blastermaster:
If you would like to only generate new QNIFFT electrostatic grids, you do not need to run blastermaster. Note, you do need the following input files generated from blastermaster:


-box (we use this to trim down the electrostatics grid to the size of the van der Waals and ligand desolvation grids)
- box (we use this to trim down the electrostatics grid to the size of the van der Waals and ligand desolvation grids)
-qnifft.parm (this is the input file for QNIFFT)
- qnifft.parm (this is the input file for QNIFFT)
-amb.crg.oxt (the charge file necessary for QNIFFT in which you tarted certain atoms)
- amb.crg.oxt (the charge file necessary for QNIFFT in which you tarted certain atoms)
-vdw.siz (radius file for QNIFFT calculation)
- vdw.siz (radius file for QNIFFT calculation)
-receptor.crg.lowdielectric.pdb (the pdb file including your receptor and low dielectric spheres)
- receptor.crg.lowdielectric.pdb (the pdb file including your receptor and low dielectric spheres)


Change into your working directory, and make sure these files are also inside. Then run the following commands:
Change into your working directory, and make sure these files are also inside. Then run the following commands:

Revision as of 23:36, 12 December 2017

Tarting refers to the polarization of specific atoms in the protein receptor to modify (enhance/decrease) ligand preferences for specific parts of the binding site. Generally one would redefine the partial charges distribution of a specific amino acid, and then use this modified residue when calculating the electrostatic potential for the receptor. Below is a step by step example:

  • Let's say you ran docking against RSK2, so you have a docking directory that looks like: rec.pdb, xtal-lig.pdb, dockfiles, working, INDOCK
  • Make a new dir e.g. RSK2-tart, and copy some input files into it:
    • mkdir RSK2-tart
    • cp RSK2/rec.pdb RSK2/xtal-lig.pdb RSK2-tart
    • mkdir -p RSK2-tart/working
    • cp RSK2/working/rec.crg.pdb RSK2-tart/working
    • cp RSK2/working/prot.table.ambcrg.ambH RSK2/working/amb.crg.oxt RSK2-tart
  • Now edit the files to represent your tweaked polarization. Let's say we want to polarize the backbone of MET496 (In this case the kinase hinge region) by +-0.4
    • Edit RSK2-tart/working/rec.crg.pdb to rename "MET A 496" to "MEU A 496"
    • Edit RSK2-tart/prot.table.ambcrg.ambH to add a MEU amino acid. You can copy the block for MET and rename it to MEU, to polarize the backbone, change the partial charge of the backbone amide proton by +0.4 (i.e. "H MEU 0.648" instead of "H MEU 0.248") lower the backbone carbonyl oxygen by the same quantity ("O MEU -0.500" to "O MEU -0.900")
    • Edit RSK2-tart/amb.crg.oxt to reflect the same changes
    • Note that the prot.table.ambcrg.ambH and amb.crg.oxt are not in working. they must be outside the working dirtory, the scripted blastermaster will copy them.
  • Now all that's left is to run the protein preparation script (blastermaster.py) with the modified parameter files, and without re-protonating the protein:
    • $DOCKBASE/proteins/blastermaster/blastermaster.py --addNOhydrogensflag --chargeFile=/path/to/RSK2-tart/amb.crg.oxt --vdwprottable=/path/to/RSK2-art/prot.table.ambcrg.ambH

These paths (e.g. /path/to/RSK2-tart/amb.crg.oxt) must be the full path the current directory "./" or relative paths "../../something/" will not work.

When the preparation is done, your protein should be "Tarted". You can proceed to docking.

You may want to visualize your modified potential maps. See the following page on how to do this with pymol:

Visualizing_delphi

Quick Tarting

If you would like to only generate new QNIFFT electrostatic grids, you do not need to run blastermaster. Note, you do need the following input files generated from blastermaster:

- box (we use this to trim down the electrostatics grid to the size of the van der Waals and ligand desolvation grids)
- qnifft.parm (this is the input file for QNIFFT)
- amb.crg.oxt (the charge file necessary for QNIFFT in which you tarted certain atoms)
- vdw.siz (radius file for QNIFFT calculation)
- receptor.crg.lowdielectric.pdb (the pdb file including your receptor and low dielectric spheres)

Change into your working directory, and make sure these files are also inside. Then run the following commands:

$DOCKBASE/proteins/qnifft/bin/qnifft22_193_pgf_32
$DOCKBASE/proteins/blastermaster/phiTrim.py qnifft.electrostatics.phi box trim.electrostatics.phi

Next, copy your new "trim.electrostatics.phi" grid into your "dockfiles/" folder. Make sure that your INDOCK file has the correct size ("delphi_nsize"). This is the value that was output by the phiTrim.py function.