jdavison wrote:
I noted that in AD the automatic bonds can be deleted, but I found after generating a pdbqt file, and then reading it, they are re-created.
Perhaps the relevant question is whether those bonds drawn in AD have any effect in the modeled docking. We're using vina to model that, once we have generated the receptor and small molecule pdbqt files.
That may not have a simple answer. It appears bonds are not specified in the pdbqt files so I am thinking the AD-created bonds don't have an effect in docking. Is that correct?
You are right, bonds are not specified in pdbqt files and they are not used in AD calculation. They are used in specifying torsions, however. My understanding is that if you delete a bond and then create pdbqt file, then that bond might influence how torsions are calculated in ADT. When you convert that molecule, with deleted bond, into pdbqt then read that with ADT, it computes bonds based on how far atoms are from each other. That's why whether you deleted a bond or not, you'll get the same picture. The only possible influence of deleting a bond on AD would be on how torsion tree are calculated.