Looks like there are bugs in mdread/ew_setup for pysander with igb=0. The
ntb is not passed in mdread1.F90, and thus the default ntb=1 is used for
igb=0. After correcting this, another error occurred at firstbox which
needed to read the inpcrd file. By providing an inpcrd file to it, the
pysander with igb=0 can generate the same energy as that of sander program.
However, if ibelly is turned on, the belly mask gives different matched
atoms than that of sander program and thus produce different energies.
Looks like there are still other bugs for ibelly.
On Fri, Feb 24, 2023 at 12:15 AM liu junjun <ljjlp03.gmail.com> wrote:
> Hi
>
> I have a test system which runs successfully in gas phase (igb=0) with
> sander. The input is as follows:
> -----------
> &cntrl
> imin = 1,
> maxcyc = 0,
> ntb=0,
> cut = 999.0,
> ibelly=1
> bellymask=":CU<:15&!:CU"
> /
> -----------
>
> I am trying to run the same task by using pysander with the following code:
> ---------
> import sander
> from parmed.amber.readparm import Rst7
>
> options = sander.gas_input(0)
> options.ibelly = 1
> options.bellymask = ":CU<:15&!:CU"
> options.igb=0
>
> rst = Rst7.open("pro.crd")
> sander.setup("pro.top",rst.coordinates, rst.box, options)
> e, f = sander.energy_forces()
> -----------
>
> Then I got the following error:
> -----------
> ASSERTion ' n > 0 ' failed in
> .../amber22_src/AmberTools/src/sander/ew_setup.F90 at line 315.
> -----------
>
> I also tried to run with igb=6, but it requires all the moving atoms to be
> before the fixed atoms.
>
> Any advice? Thanks a lot!
>
> Junjun
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Feb 24 2023 - 02:00:03 PST
