On Mon, Apr 24, 2017, Rakesh Srivastava wrote:
>
> I am using 3D-RISM for thermodynamics study of a protein. I use certain
> (say 10) number of frames with MPI version of RISM but due to convergence
> problem it fails at any frame in between. When I again submit the job, it
> fails at some other frame where it ran successfully last time. I don't
> understand how to control convergence in 3D-RISM. I also tried with smaller
> step size of numerical solver. If anyone could help, I will be thankful.
I generally use npropagate=5, mdiis_del=0.7, mdiis_nvec=5, tolerance=1e-8;
for just energies (as in MM-3DRISM) you may be able to get away with a
somewhat looser tolerance.
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Apr 24 2017 - 05:00:02 PDT
