Author Topic: Running same method calculation on multiple molecules.  (Read 35 times)

prasanta13

  • Full Member
  • ***
  • Posts: 28
  • Karma: +0/-0
Running same method calculation on multiple molecules.
« on: June 13, 2022, 07:50:36 AM »
Hi,
I want to run PBE/def2-TZVP calculation on 1200 molecules. Now it can be done with Tmolex, but I have a problem with the graphic driver (probably) and it doesn't work as expected. Starting define module for each system is not a way. Is there any other way available so I can generate the input and run Turbomole in one go?

uwe

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 511
  • Karma: +0/-0
Re: Running same method calculation on multiple molecules.
« Reply #1 on: June 13, 2022, 08:15:01 PM »
Hi,

sure, there are quite a number of options which all allow running calculations on lists of molecules. But before I start to list those, it seems that you would prefer to use TmoleX.

So let me first ask why you want to run TmoleX on a remote machine at all. While there are remote desktop solutions like RDP, VNC, Nomachine, Teamviewer or even the good old X11 server still works, it is highly recommended to install and use the graphical user interface locally on the machine you are using as a desktop. That guarantees a responsive user interface.

Is the machine you want to run the calculation on accessible by secure shell (ssh)? If yes, you can use the Run(network) button in TmoleX instead of the Run(Local) one when starting jobs. It asks you for the name of the remote machine, user name, password and the installation directory of the command line version of Turbomole on that machine. If you have all that, you can run jobs the same way you are used to on your local machine.
Queuing systems like PBS, SLURM, LSF, etc. are supported and jobs can be submitted to those from TmoleX too.

Would that be an option?


prasanta13

  • Full Member
  • ***
  • Posts: 28
  • Karma: +0/-0
Re: Running same method calculation on multiple molecules.
« Reply #2 on: June 14, 2022, 01:17:16 PM »
Sure, Tmolex offers an excellent choice. But there are only a few issues.
1. The program crashes if large number of molecules are present.

2. The choice of methods (basis set and other properties) are limited in Tmolex.

I used Tmolex extensively for few purposes, single-point calculation etc. However, I need to do calculations with aug-cc-pVQZ (Tmolex only has def2 basis).