On Wed, Aug 08, 2018, John L. Kulp, Jr. wrote:
> Is it helpful to compile (on linux) with architecture-specific flags
> and/or openmp to speed up sqm? This is for doing AM1 calculations (as
> part of getting am1-bcc charges).
I don't think such tweaks are likely to be helpful.
>
> Does it work to use PM6 for getting initial charges for am1-bcc, or is
> the parameterization only for AM1? The reason is to get better
> convergence in some cases.
The parameterization is for AM1 only. There are some notes about
helping convergence on p. 287 of the Amber 2018 Reference Manual.
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Aug 08 2018 - 12:00:04 PDT
