some thoughts - others on the list may also have suggestions
1) bubbles shouldn't form during minimization, things should not move far.
They do form during MD.
2) you might want to reduce the closeness on your solvateoct from 1.0 to
maybe 0.8 (we use 0.8)
3) 500 is a very strong force constant and may cause problems on it's own.
try 10 unless you're sure you need this.
4) your equilibration should include a constant pressure step early on,
likely with the MC barostat. Look at Dan Roe's paper on equilibration
protocols.
On Wed, Feb 2, 2022 at 1:06 PM Kenneth Huang <khuang8.student.gsu.edu>
wrote:
> Hi all,
>
> I'm having an unusual situation where I have a single nucleoside (DTN in
> this case) in a water box with salt on Amber16, but it keeps crashing
> during production. I went back and double checked, and noticed there was
> some very odd behavior with void bubbles forming during the initial
> minimization (attached), ie if I run it with
>
> mpirun --bind-to core -np 6 ${AMBERHOME}/amber16/bin/pmemd.MPI -O -i
> 01a_sol_min.in -p *.prmtop -c *.inpcrd -o 01a_sol_min.out -r
> 01a_sol_min.rst -ref *.inpcrd
> &cntrl
> imin = 1,
> maxcyc = 2000,
> ncyc = 1000,
> ntmin = 1,
> ntb = 1,
> ntr = 1,
> ntpr = 500,
> cut = 8.0
> /
> Hold only DNA+protein fixed
> 500.0
> RES 1 1
> END
> END
> Some very large gaps form in the solvent which by the point it hits
> production I'm guessing it can't resolve, but I can't think of why that'd
> be? I did a fairly standard setup for the system with
>
> source leaprc.DNA.OL15
> source leaprc.water.tip3p
> loadamberparams frcmod.ionsjc_tip3p
> m1 = loadpdb DTN_21.pdb
> solvateOct m1 TIP3PBOX 10.0 iso 1.0
> addIons m1 Na+ 0
> addIons m1 Cl- 0
> addIonsRand m1 Na+ 2
> addIonsRand m1 Cl- 2
> saveamberparm m1 DTN_21.prmtop DTN_21.inpcrd
> savepdb m1 check_DTN_21.pdb
> quit
>
> I've managed to get other nucleosides to finish without any issues so far,
> but I'm scratching my head of what'd be wrong, or if I had missed something
> obvious?
>
> Best,
>
> Kenneth
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Feb 02 2022 - 12:00:02 PST
