Project

General

Profile

Bug #810

CHARMM cp=0 dihedrals are not ignored

Added by Roland Schulz over 7 years ago. Updated over 6 years ago.

Status:
Closed
Priority:
Low
Assignee:
Category:
-
Target version:
Affected version - extra info:
Affected version:
Difficulty:
uncategorized
Close

Description

I noticed that the CHARMM FF has dihedrals defined with the force constant cp=0. And they don't seem to be automatically removed.

Erik added the comment to assign_parm (conparm.c): * Second, if the force constant is zero in both A and B states, we set the phase * and multiplicity to zero too so the interaction gets removed during clean-up.

But I cannot find any code which does this clean-up (neither in grompp nor mdrun) and I also just tested that the cpA(=cpB)=0 functype is in the tpr file.

Am I missing something or are the cp=0 dihedrals not removed? If so where should they be removed?

Associated revisions

Revision 74bb04c1 (diff)
Added by Erik Lindahl over 6 years ago

Add check to remove zero Charmm dihedrals

Proper torsions where the force constant is zero
in both A and B states are now removed. We also
check for other angle, torsion, and restraint
functional types, and if all parameters are zero
for these the interaction is not added. This will
not change any results, but increase performance
slightly by not calculating unnecessary interactions.
Fixes #810.

Change-Id: I37ecd06d0641008593edab29e5b08433bde7b6cc

History

#1 Updated by Erik Lindahl over 7 years ago

  • Priority changed from Normal to Low
  • Target version changed from 4.5.5 to 5.0

The removal used to be done while converting parameters.

However, I would classify this as a hard bug since it doesn't affect the quality of any results (actually, it won't affect results at all, but we end up computing a couple of zero interactions). We should look into it some time, but there's no reason to hold 4.5.5 for this.

#2 Updated by Erik Lindahl over 7 years ago

would -> wouldn't

#3 Updated by Roland Schulz over 7 years ago

For my system it is performance relevant. So I would suggest to fix it for 4.6. I'm happy to help with it but I would need to know where we want to filter them.

#4 Updated by Roland Schulz over 7 years ago

  • Target version changed from 5.0 to 4.6

#5 Updated by Rossen Apostolov about 7 years ago

  • Assignee changed from Pär Bjelkmar to Erik Lindahl

#6 Updated by Mark Abraham almost 7 years ago

  • Assignee changed from Erik Lindahl to Mark Abraham

Can you contribute a test case, please Roland?

#7 Updated by Roland Schulz almost 7 years ago

You can use standard 1aki.
pdb2gmx -f 1AKI.pdb -ff charmm27 -water tip3p; touch empty.mdp; grompp -f empty.mdp ; gmxdump -s topol.tpr|less
functype 884 has cpA = 0.

#8 Updated by Erik Lindahl over 6 years ago

  • Status changed from New to In Progress

#9 Updated by Erik Lindahl over 6 years ago

  • Status changed from In Progress to Closed

Also available in: Atom PDF