# Ligand topology (CGenFF) with high charge penalties

**URL:** <https://gromacs.bioexcel.eu/t/ligand-topology-cgenff-with-high-charge-penalties/5423>\
**Category:** User discussions\
**Created:** [December 12, 2022, 5:00pm UTC](https://gromacs.bioexcel.eu/t/ligand-topology-cgenff-with-high-charge-penalties/5423 "2022-12-12T17:00:16Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![amadafghilardi](https://avatars.discourse-cdn.com/v4/letter/a/8c91f0/32.png) [@amadafghilardi](https://gromacs.bioexcel.eu/u/amadafghilardi)\
**Post date:** [December 12, 2022, 5:00pm UTC](https://gromacs.bioexcel.eu/t/ligand-topology-cgenff-with-high-charge-penalties/5423/1 "2022-12-12T17:00:16Z")

</div>

GROMACS version: 2022.2  
GROMACS modification: Yes

I am trying to use MD to evaluate a docking pose obtained with autodock vina. But I am having issues generating the ligand topology using the CGenFF server. I followed all the same steps shown in the tutorial (Protein-Ligand complex). I converted the ligand to mol2, and fixed the bonds’ order with the script “sort\_mol2\_bonds.pl” provided on the GROMACS page. The ligand\_fix.mol2 file seems correct when opened in ChimeraX and contains all hydrogens. However, the .str file generated contains high charge penalties and the pdb resultant does not show the proper structure anymore. What is the possible reason and how could I fix this issue? Thanks!

---

<div class="post-metadata">

**Author:** ![jalemkul](https://dub1.discourse-cdn.com/flex017/user_avatar/gromacs.bioexcel.eu/jalemkul/32/18_2.png) [@jalemkul](https://gromacs.bioexcel.eu/u/jalemkul)\
**Post date:** [December 12, 2022, 6:58pm UTC](https://gromacs.bioexcel.eu/t/ligand-topology-cgenff-with-high-charge-penalties/5423/2 "2022-12-12T18:58:16Z")

</div>

> [@amadafghilardi](#):
>
> However, the .str file generated contains high charge penalties and the pdb resultant does not show the proper structure anymore.

Is your residue name 4 characters or shorter in the `.str` file (RESI XXXX)? If not, you’ll get a misformatted PDB file.

As for the high penalties, check the CGenFF papers (particularly the original 2010 JCC paper) for how to validate and refine a topology. It will require QM calculations and is most easily done in the CHARMM software (for some things, there are no equivalent functions in GROMACS).