My ligand has a quaternized pyridine ring in it's structure, but after editing, when I export it from AutoDock Tools as a .pdbqt file and open it in Discovery Studio Visualizer, the structure of my ligand changes and nitrogen appears to no longer be quaternized (a C=N double bond is lost, the +1 charge on the nitrogen is lost, there are now only 2 double bonds in the ring). The rest of the ligand stays as is, the problem appears to be happening only to the quaternized nitrogen. Is there a way to fix this from happening?
The pdb, and as an extension the pdbqt format, do not encode the bond order. These are very old coordinate files originally designed as memory-saving formats for for macromolecules, and rely on the interpreting program having an internal dictionary of the different residues than make up a protein or nucleic acid. They are not very suitable for conveying the detailed informations for small molecules from one program to the other - this is why you have to supplement the coordinate files with specific parameter files describing small molecules and non-standard residues for energy minimisation and molecular dynamics programs - these expand the internal residue dictionaries with the description of those non-standard building blocks.
Dear Vice
as Annemarie pointed out, Autodock actually convert the dative bonds (e.g. [N+]([O-])=O to N(=O)=O).
Have you tried obabel ?
use
obabel input.mol2 -O output.pdbqt
or
obabel input.sdf -O output.pdbqt --gen3d
Tanuj Sharma Thank you for your answer! Yes, I have tried using obabel; when I convert my ligand file from .sdf of .mol to .pdbqt and view it in PyMol or DSV, the same problem is still present - one double bond missing.
Of course, the structures (two double bonds, a positive and a negative charge on opposite sides or three double and three single bonds alternating around the ring) are resonance structures all encompassed by the aromaticity of the ring. To see the actual information your program is using, you have to look at positions 67 - 76 in the pdbqt file, these denote the partial charges assigned to the individual atoms in the ring. Calculations will be based on these values, the graphical representation is quite independent of that.
Annemarie Honegger Thanks again for taking the time to answer my question! You're right; when I view my edited ligand.pdbqt file in PyMol, all double bonds seem to be present, but they are missing when I open the same file in Discovery Studio. Since they don't show up in DS, do you think I could just use my .pdbqt file as is, run Vina, get my ligand poses and binding affinity results, then, in Discovery Studio, just edit my ligands and add the "missing" double bond and +1 charge on the nitrogen and dock the ligands. Or would that result in inacurrate poses, binding affinites and receptor-ligand interactions? I'm still really new to this and I'm very curious to hear what you think!
Yes. The programs/subroutines that do the calculations do this based on the properties of the individual atoms, independent of the bond model. There is no localised +1 charge on the nitrogen, which you would have in the protonated nitrogen, but the lone electron pair of the nitrogen is part of the aromatic pi electron system.
- Badges
- Science topic
- Similar topics
- Filing