I want to restrict which residues my small molecules will interact with. Several of the near by residues can be variable. However I can not restrict the search grid to this area because the orientation does not line up.
I wonder why you want it to do that unusual way. Although i never tried it but still a possible solutions for this may be removal of those amino acid residues from your basic input PDB file of protein. This will simply cut out them from docking calculations and your grid box will remain intact.
This would allow improper docking. Small molecules would be sit in the space these residues take up. I need to be able to disregard poses that interact with the spatial coordinates of these residues.
Yes you are true and i understand that but after docking you may superimpose the both protein file and look for poses of docked ligand in which ligand is not occupying those 3D spaces reserved for resides not of interest.
Depending of each kind of interactions you want to switch off with these side chains, you can set to zero the force field parameters to zero. For instance, if the electrostatic interactions are to turn off, set the partial atomic charges to zero.
Ranjnikant: I am screening several thousand small molecules. Is there a way to have a program determine whether or not there is overlap of protein with small molecules?
Hiqmet: At the moment I am using autodock vina because of its enhanced speed. This program does not use user specified forcefield parameters. What program would let me switch off the side chains?
First, you should see the files you prepare as setup for the docking software, I think they are files with extension .pdbqr, which contains the charges and vdW radius of atoms. Here, you can modify either the charges or vdW radius of these atoms you want. If you use Vina with Autodock tools and PyMol (in MGLTools) you may be able to create these files and modify.
In principle, you can also modify the force field topology files, by renaming the residues that you want to change. These files should be in the software installation directory.
Of-course. You can develop your own script to screen such poses by examining the matching 3D coordinates of excised out residues and final docked pose of ligand. Adopting exact same coordinates combination will be rare and obvious but you can use average and threshold screening for such purpose. A mix of shell and PERL scripts might lead you towards filtered poses.
- Badges
- Science topic