pbm running ISOLDE (OpenMM)

[vincent.chaptal@…]

Hi,

I have an error message trying the run the plugin ISOLDE coming back 
from vacation. It worked before leaving...

I load a pdb just fine, load the map fine, initialize fine, but when I 
hit the run button from ISOLDE I have this:
the error message is:
File 
"/Applications/ChimeraX.app/Contents/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/simtk/openmm/app/forcefield.py", 
line 1051, in createSystem
template = self._templates[tname]

/See log for complete Python traceback.

/looking at the forcefield.py, the template = self._template line is 
under the "unmatched residues" part of the program.

Does it mean that there is something going on with my pdb with unmatched 
residues between chains (4 identical chains), or is it more a "computer 
problem"?
something that might help is that I recall updating my OS system (mac 
10.11.6), but I'm not sure of the dates I've done it and if I used 
ISOLDe afterwards or not.

Thanks for your help
Vincent
//

-- 

Vincent Chaptal, PhD

Institut de Biologie et Chimie des Protéines

Drug Resistance and Membrane Proteins Laboratory

7 passage du Vercors

69007 LYON

FRANCE

+33 4 37 65 29 01

http://www.ibcp.fr

[tic20@…]

Hi Vincent,

If you could send me the full traceback to check it would be great, but from what you’ve shared I think this is most likely a known bug in ISOLDE that occurs if you have any truncated CYS residues. The short story is that because of how the simulations are created, the molecular dynamics engine needs to be explicitly told about each cysteine - whether it’s normal, deprotonated or part of a disulphide. When writing the code for that I failed to consider truncations (which are treated as ALA). It’ll be fixed in the next release, but for now just make sure all your CYS sidechains are built.
 

 

[vincent.chaptal@…]

Hi Tristan,

thank you for your email.
I would send you the traceback but I don't know where to find it...

Following your guidelines, I managed to fix the problem and am running 
ISOLDE now.
this is what happened:
- I usually add hydrogens in coot unsing the refmac tool inbedded within 
coot. But this time it wouldn't work. I thus turned to pymol to add 
hydrogens. It gave me the error message pasted bellow in my original post.
- the cysteines were not truncated, all protonated. I removed the h1 on 
the sulfur atom for a test, and the error message disappeared, but was 
replaced by another error message telling me that the first LYS of the 
PDB was lacking an hydrogen. it didn't recognize the fact that it was 
LYS 9 and therefore not the first amino-acid of the protein I guess. 
Sorry I didn't keep the exact error message.
- I then restarted coot from the terminal this time, using the latest 
coot. This time hydrogens were added immediately and it worked in isolde 
right away.

Hoping this could help you.
Best
Vincent

On 27/08/2018 23:32, ChimeraX wrote:

[tic20@…]

When an error is thrown, the full traceback should appear in ChimeraX's 
Log window (on the right).

In general, the easiest way to prepare your model for ISOLDE is using 
ChimeraX's own AddH tool. On the ChimeraX command line:

delete H # optional - only if you suspect that there are erroneous 
hydrogens present
addh hbond false

AddH does a few things that are particularly useful for ISOLDE - in 
particular, in the way it handles truncated sidechains. I'd highly 
recommend using it. Meanwhile, I'm working on making sure errors from 
the MD engine are parsed and turned into something more informative.

Cheers,

Tristan

On 2018-08-28 12:49, ChimeraX wrote:

[Tristan Croll]

Just cleaning up old tickets!