Skip to content
This repository has been archived by the owner on Feb 15, 2022. It is now read-only.

command-line only error: AssertionError: PROPOSED CHARGE IS NOT POSSIBLE: SOLUTE MAY BE AN OPEN SHELL #79

Open
therealchrisneale opened this issue Jul 14, 2021 · 0 comments

Comments

@therealchrisneale
Copy link

  • I believe this to be either a bug with LigParGen (command-line version only), or perhaps the online webserver is doing additional preparatory steps that I have not included for the command-line version.

Issue Information

Software name & Version :
$ LigParGen -h
usage: LigParGen [-h] [-r RESNAME] [-s SMILES] [-m MOL] [-p PDB]
[-o {0,1,2,3}] [-c {0,-1,1,-2,2}] [-l]

Ligand Parameter Generator Based on 
Jorgensen group's OPLS-AA/CM1A(-LBCC) FF
Created on Mon Feb 15 15:40:05 2016
@author: Leela S. Dodda [email protected]
@author: William L. Jorgensen Lab 

Expected Behavior

Able to build (or not) the same chemicals with the webserver and the command-line version of LigParGen

Actual Behavior

Hello,

I find that a neutral molecule (here the lipid POPC) build with both the online web server at http://zarbi.chem.yale.edu/ligpargen/ and also with the command-line tool. However, some negatively charged molecules (here the lipid POPA) only build with the online web server and throw an error with the command-line tool.

##For POPC:
$ smiles="CCCCCCCCCCCCCCCC(=O)OCC(COP(=O)OCCN+(C)C)OC(=O)CCCCCCCC=CCCCCCCCC"
$ charge=0
$ LigParGen -s $smiles -r UNK -c $charge -o 0 -l

LBCC converter is activated
1 molecule converted
262 audit log messages
MOLECULE HAS A CHARGE of 0
Changing OS TO OE
Changing OS TO OE
Changing OS TO OE
DONE WITH OPENMM
DONE WITH Q
DONE WITH XPLOR
DONE WITH CHARMM/NAMD
DONE WITH GROMACS
DONE WITH LAMMPS
DONE WITH DESMOND

For POPA:

$ smiles="CCCCCCCCCCCCCCCC(=O)OCC(COP([O-])=O)OC(=O)CCCCCCCC=CCCCCCCCC"
$ charge=-1
$ LigParGen -s $smiles -r UNK -c $charge -o 0 -l

1.14*CM1A-LBCC is only available for neutral molecules
Assigning unscaled CM1A charges
1 molecule converted
238 audit log messages
MOLECULE HAS A CHARGE of -1
Traceback (most recent call last):
File "/usr/projects/mrmdesign/MCMD/CONDA_ENVS/my-ligpargen-env/bin/LigParGen", line 11, in
sys.exit(main())
File "/usr/projects/mrmdesign/MCMD/CONDA_ENVS/my-ligpargen-env/lib/python3.6/site-packages/LigParGen/Converter.py", line 90, in main
convert(**vars(args))
File "/usr/projects/mrmdesign/MCMD/CONDA_ENVS/my-ligpargen-env/lib/python3.6/site-packages/LigParGen/Converter.py", line 157, in convert
charge), "PROPOSED CHARGE IS NOT POSSIBLE: SOLUTE MAY BE AN OPEN SHELL"
AssertionError: PROPOSED CHARGE IS NOT POSSIBLE: SOLUTE MAY BE AN OPEN SHELL

PS: I get the same behavior if I build a PDB first with RdKit and use that as input to LigParGen instead of direct smiles string input.

Also please note that I tried both with the smiles string from a google search that gave me a smiles from abcam.com and also a cannonical version from this tool https://cactus.nci.nih.gov/translate/ and I got the same error.

Finally, I do realize that I don’t need to use LigParGen to get parameters for lipids. This is intended as an easily verifiable test case for something with a verifiably correct smiles string. My actual problems are with different phosphate containing molecules.

Thank you,
Chris.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant