On Tue, Oct 04, 2022, Rafał Madaj via AMBER wrote:
>
>Upon running MMGBSA for protein-ligand complex with Mn2+ ions present
>the script fails yielding:
>Beginning GB calculations with /opt/apps/amber20/bin/sander
> calculating complex contribution...
> bad atom type: Mn
The GB parameterizations in Amber don't know about ions. I'm not sure how
the program is looking for atom type "Mn", since the atom type from ParmEd
(below) is Mn2+.
>> printDetails :350
>
>The mask :350 matches 1 atoms:
>
> ATOM RES RESNAME NAME TYPE At.# LJ Radius LJ
>Depth Mass Charge GB Radius GB Screen
> 5669 350 MN MN Mn2+ 25 1.4010 0.0159
>54.9400 2.0000 1.5000 0.8000
You could try to "trick" the codes into using the GB Radius from ParmEd, but
one doesn't expect GB results for divalent ions to have any real accuracy.
Can you not strip the Mn2+ ions from all of your input prmtops? You might
have better luck trying to use PB rather than GB, but I doubt it. If you
are stripping waters but not Mn2+, you will end up with a divalent ion with
no first-shell ligands, and this is unlikely to fare very well in energy
calculations.
Others on the list may have more specific suggestions.
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Oct 04 2022 - 12:00:03 PDT
