On Thu, Oct 29, 2020, h_sahakyan.mb.sci.am wrote:
>I would like to perform MM/GBSA for a protein-ligand complex after
>minimization with an implicit solvent. In this regards, I have several
>questions:
>1) What kind of radii is better to use for this task and is there any
>minimal recommended cutoff for implicit solvent simulations?
Note that MM/GBSA is intended to be used with snapshots from MD, not
from minimization. Best recommendation for parameters is igb8 and
mbondi3. See p. 70 in the Amber 2020 Reference Manual for more
information.
>
>2) What should be the ratio of steepest descent and conjugate gradient
>cycles (maxcy, ncyc)?
I recommend using ntmin=3 for any serious minimization.
>3) I noticed that it is recommended to perform the minimization on CPU,
>what kind of consequences can have minimization on GPU with pmemd.cuda
The GPU version does not use any cutoffs for nonbonded interactions,
which is probably good for things like MM/GBSA.
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Oct 30 2020 - 05:00:02 PDT
