Previous Thread
Next Thread
Print Thread
Ringed structures with poor dihedrals at CGenFF
#37851 04/23/20 03:11 PM
Joined: Apr 2020
Posts: 2
A
ashim Offline OP
Forum Member
OP Offline
Forum Member
A
Joined: Apr 2020
Posts: 2
I have a small molecule, 4-aminopthalimide with a purely aliphatic group attached to its terminal end, that I want to use for MD in a CHARMM-GUI made lipid membrane, using GROMACS.

I optimised the structure on HF/6-31G* using ORCA, and then used the CGenFF server to get the Charmm36 compatible .str file during which I get ~20 penalty values on charges, and high (>50) penalty values on dihedrals of the ringed structure.

I tried using the ffparam-gui (with psi4 for QM and OpenMM for MM) to optimise the parameters. I am unable to optimise the dihedrals, due to some issues arising from the bonds to be parametrized being part of the ring preventing psi4 from being able to run PES scans.

I have attached the initial .str file as obtained from the CGenFF server, as well as the .mol2 file of my molecule.
4apC9.mol2
4apc9.str

Since the majority of the issues seem concentrated in the ringed structure, I tried to parametrize the molecule without the aliphatic chain. While charge parametrization seems to have worked, the dihedral issue still remains. (Also, is there any way to validate these charges?)

Is there a better way to go about this? Kindly help. (I don't have access to Gaussian or a CHARMM license, yet.)

Regards
Ashim

Re: Ringed structures with poor dihedrals at CGenFF
ashim #37852 04/23/20 09:28 PM
Joined: Sep 2003
Posts: 8,498
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,498
These files would be easier to view if they had been attached (or referenced) as .txt files, as outlined in the READ BEFORE POSTING topic.

The high penalties do not mean the parameters are bad, just that the five membered ring topology is not well represented in the database.

This would appear to be a planar fused ring aromatic system, so I would run some gas phase low friction Langevin simulations of the new molecule, and a few similar molecules already in the CGenFF topology file, such as pyrimidines and other fused ring aromatic molecules. Then, align the trajectory coordinates using the ring heavy atoms, compute the average coordinates, and finally compute the fluctuations about the average. If the ring atom fluctuations are comparable to those from existing known planar molecules, that would be reasonable first pass test.


Rick Venable
computational chemist

Re: Ringed structures with poor dihedrals at CGenFF
rmv #37854 04/24/20 05:19 AM
Joined: Apr 2020
Posts: 2
A
ashim Offline OP
Forum Member
OP Offline
Forum Member
A
Joined: Apr 2020
Posts: 2
Apologies for the format issue. I'll try what you wrote, (I'm new to this) and get back here.

Last edited by ashim; 04/24/20 05:20 AM.
Re: Ringed structures with poor dihedrals at CGenFF
ashim #37855 04/24/20 03:12 PM
Joined: Sep 2003
Posts: 8,498
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,498
I should note that the CGenFF program results are dependent on the correct atom typing in the .mol2 file; high penalties can also be the result of atom typing issues.

Web browsers rely on filename extensions like .html, .png, or .txt to know how to display a file, and most do not know what to do with a file that ends with .str or .mol2


Rick Venable
computational chemist


Moderated by  alex, lennart, rmv 

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.4
(Release build 20200307)
Responsive Width:

PHP: 5.6.33-0+deb8u1 Page Time: 0.013s Queries: 22 (0.006s) Memory: 0.9135 MB (Peak: 0.9952 MB) Data Comp: Off Server Time: 2020-09-25 13:08:26 UTC
Valid HTML 5 and Valid CSS