Closed kexul closed 3 years ago
I've consulted an expert, who told me that the .top
file generated by protocaller had some missing dihedrals parameters.
Besides, the dummy atom's name seems to be duplicated with previous atoms, would that be a problem ?
It does seem that there are issues with the dummy atoms but it's not obvious what the nature of these is. I think the best course of action is to try the opposite perturbation which should result in unique atom names and try running these. It is also worth comparing the dihedrals from each separate ligand and make sure that they are correctly represented in the merged topology, but it is also possible that GAFF doesn't return any dihedral terms for this particular bond, so it is set to 0.
I just looked at your mdp
file. You are turning on the electrostatic interactions for ligand B before turning on the Lennard-Jones interactions, and this is most likely the reason for the behaviour you are observing.
I just looked at your
mdp
file. You are turning on the electrostatic interactions for ligand B before turning on the Lennard-Jones interactions, and this is most likely the reason for the behaviour you are observing.
Wow! I changed the order as you said, which did the trick. Thanks so much for your advice! By the way, the old order I followed was taken from the example of ProtoCaller, except increasing the number of lambda windows. Would you mind elaborate the difference between the cases? In which situation should we turning on the LJ before electrostatic?
@kexul you are changing a small molecule to a larger one. Therefore you need to turn on the LJ before electrostatic; otherwise, charges will appear in the empty space which will lead to instabilities of the simulation.
@tsenapathi Got it, Thank you!
Hi, here is my transformation
here is the
.gro
and.top
file generate by ProtoCaller, morphs.zipThe minimization step seems fine, but, when running NVT, there are many LINCS warning, below are some of them:
When I check the gro file generate after NVT Equilibration, atom
49
and atom43
are too close, which I think should not happen in reality.Here is the
.mdp
file I used and.gro
file generated after NVT Equilibration. files.zipAny clue what happened? Thanks in advance!