Ligands prepped with Maestro

[adw62]

I prepared two ligands with Schrodinger Maestro they seem to be causing webTIES some problem.

Issue maybe be with the ligands I have not verified them elsewhere: lig.zip

[bieniekmateusz]

Just replied on the email. For now it's best to remove the ions, and later we can account for it. Please feel free to set the priority here. Cheers

[bieniekmateusz]

I just run in the command line

ties create -l l1.mol2 l2.mol2 -nc -1 -uniq-a true

The error in parmchk2, I'll check later to sort it out. Atom type of C.ar does not shown up in PARMCHK.DAT

This is probably because .mol2 by default is trying to reuse the charges and atoms types.

[bieniekmateusz]

I've updated the server now and it should now recognise that the .mol2 can have charge=0 used as placeholders, meaning that BCC has to be computed. That is now being done as a job for a worker. Unfortunately, this job by default uses many cores, meaning that it steals resources from other workers (one worker per core). Ideally I'd be able to tell antechamber to use just one core for that (even if that means that it is going to take more time per case).

Anyway, let me know how that is working.

[adw62]

That is all working for me now thanks :) I agree with you comments on resources, might get expensive if many people compute bcc, I can't see anything immediately in the AMBER manual about cores. Maybe something like this would work: https://unix.stackexchange.com/questions/23106/how-to-limit-a-process-to-one-cpu-core-in-linux?

Could do something like:

subprocess.run(['taskset', '-c',  '0', self.config.ambertools_antechamber,
                          '-i', self.current, '-fi', filetype,
                          '-o', new_current, '-fo', 'mol2',
                          '-dr', self.config.antechamber_dr],
                         stdout=LOG, stderr=LOG,
                         check=True, text=True,
                         cwd=cwd, timeout=30)

Edit: Thinking about this more it doesn't work because everything will get pinned to CPU 0

[bieniekmateusz]

Thanks, this looks promising, I'll give it a go, thanks!

[adw62]

Think this was fixed