Re: [AMBER] generated parameters in leap

From: Pengfei Li <>
Date: Tue, 9 Jul 2019 22:45:33 -0400

Hi Sarah,

The strategy of renaming some atom types was correct but this process can be complicated due to so many new atom types you have now.

One suggestion is to use the “sed" command to rename the atom types in the second and third of the generated frcmod files (e.g. replacing M1 by M2, etc) to prevent parameter conflicts and human errors, and do not combine these frcmod files but read them separately in the final LEaP input file.

Hope it helps,

> On Jul 5, 2019, at 1:13 PM, SARAH JEANNE LEFAVE <> wrote:
> Hi all,
> I am parameterizing a protein with 3 different metal centers (two heme sites with different charges on the metal and substituents, and a Cu histidine site) I have finished steps 1-4 and am working on generating the parameter and topology files. I ran 3 separate runs for each of the metal sites as they are distinct active sites from one another. This generated 3 different sets of files for running with tleap. I combined these into one adding the additional atoms to be defined and loading all the relevant files in. After getting many errors about **no angle parameters and ** not torsion terms, I redefined the redundant metals and nitrogen’s that were added in the addAtomTypes section such that I now have (M1-M3 and Y1-14). I also modified the mol2 and frcmod files to be consistent for these. While this reduced some of the errors regarding angle parameters and torsion terms I still get a significant amount. I suspect it is due to the redundant cc, ce, cd, ha… and CR, CV, HX… etc. If this is in fact the problem is there another way to change this rather than manually renaming the atoms in the ligand? Will renaming them to non-redunant names cause any other error for recognition with gaff or otherwise?
> Thank you!
> Sarah
> _______________________________________________
> AMBER mailing list

AMBER mailing list
Received on Tue Jul 09 2019 - 20:00:02 PDT
Custom Search