search

selimsami / qforce

Apache License 2.0
57 stars 13 forks source link

Add torsiondrive as a potential torsion scan tool. #43

Closed xiki-tempula closed 2 years ago

xiki-tempula commented 2 years ago

@selimsami Do you mind help me on this? I have allowed the torsiondrive to be used as torsion scan tool but it seems that the charge is not read in correctly.

https://github.com/Exscientia/qforce/blob/0fded8348d59a9967048f525c2fb766580162669/qforce/qm/qm.py#L122 I'm loading the charge from the torsion scan to be used as charge for the fitting.

But I found that the charge that I feed in to the Qforce (propane_qforce/fragments/CC_H8C3_d91b46644317dee9c2b868166c66a18c~1.charges is not the same as the topology that Qforce generates (Qforce (propane_qforce/fragments/CC_H8C3_d91b46644317dee9c2b868166c66a18c~1/step00/propane_qforce.itp) during the fitting.

Please find the example. Archive.zip

xiki-tempula commented 2 years ago

@selimsami Sorry, my fault. Forget to set

[ff]
charge_scaling = 1.0
xiki-tempula commented 2 years ago

@selimsami The code is done now and I'm testing the robustness. Do you mind having a look and tell me your thought, please? Thanks. Sorry that I need to change the sys.exit to raise SystemExit as otherwise my pycharm won't work https://youtrack.jetbrains.com/issue/PY-53625. Sorry, I have to use some bash script again as torsiondrive command is a bit complicated and the output is an xyz file.

xiki-tempula commented 2 years ago

@selimsami Thanks for the advice on ase. Once I have tested that it is robust, I will ping you and we can have this merged.

xiki-tempula commented 2 years ago

@selimsami I have tested this routine on 30 molecules and it seems fine. If you are happy, we can have this merged.