Re: [AMBER] Non-convergence when calculating partial charges for aromatic-adjacent nitriles

From: David A Case <>
Date: Mon, 5 Nov 2018 14:11:22 -0500

On Fri, Nov 02, 2018, Jeffrey Wagner wrote:
> I am attempting to run partial charge calculations for a set of molecules.
> I've found that sqm fails frequently on nitriles connected to aromatic
> systems.

Please try the suggestion on p. 287, Note 6, (especially adding
"ndiis_attempts=700,"; I'd suggest leaving the scf convergence unchanged).
For me, your file runs fine with this change.

This really should be the default, but it would break backwards
compatibilty and might make some simpler molecules go slower, so
we've never done it.

(Developers: one possibility is to intercept an SCF convergence error,
and re-run with the option above.)


AMBER mailing list
Received on Mon Nov 05 2018 - 11:30:02 PST
Custom Search