On Wed, Sep 28, 2016 at 4:46 PM, Elisa Pieri <elisa.pieri90.gmail.com>
wrote:
> Dear all,
>
> I have a PDB containing a protein inserted in a membrane, plus two water
> layers, yielding a "cubic" unit. Here the tleap input I used to get the
> parameters:
>
> source leaprc.constphRET (<-- a customized leaprc for the cromophore)
> source leaprc.lipid14
> loadAmberParams frcmod.ionsjc_tip3p
> monomer = loadPDB monomer.pdb
> setBox monomer "vdw"
> saveAmberParm monomer monomer.parm7 monomer.rst7
> quit
>
> Now..I needed to minimize, so I used this input:
>
>
> &cntrl
> imin = 1, ! Turn on minimization
> igb = 2, ! Use the GB model CpHMD was parametrized
> for
> saltcon = 0.1, ! Use the salt conc. CpHMD was parametrized
> for
> maxcyc = 5000, ! Total number of minimization cycles
> icnstph = 1, ! Turn on constant pH
> ntcnstph = 100000, ! Never attempt to change prot. states
> cut = 1000.0, ! No cutoff
> ntr = 1, ! Turn on positional restraints
> restraint_wt = 10, ! 10 kcal/mol/A**2 restraint force constant
> restraintmask = '!:WAT&@CA,C,O,N' ! Restraints on the backbone atoms
> only
> ntxo = 1
> /
>
You shouldn't use a constant pH implicit solvent mdin file for an explicit
solvent simulation. Weird things will happen. If you minimize without PBC
and then try to heat or do other things *with* PBC, you will almost
certainly have problems with particles that "collide" with periodic images
(that were not modeled in the minimization).
You should use the same potential energy function for all steps of the
simulation. You can find a template explicit solvent mdin file for
minimization, heating, etc. at one of my GitHub repositories:
https://github.com/swails/Mdins
There is a description of each mdin file and what it's for in the readme.
> So..second question: is this happening because of my "homemade periodic
> conditions solution", or the heat.mdin is simply wrong somewhere?
> Here the specific of what I'm doing: I'm trying to prepare my system for
> CpHMD in explicit solvent, using the CUDA code of Amber16. Can you help me?
>
I think it's your homemade PBC workflow. Have you looked at my CpHMD
explicit solvent tutorial?
http://jswails.wikidot.com/explicit-solvent-constant-ph-md That should help
you get started, I think.
HTH,
Jason
--
Jason M. Swails
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Sep 29 2016 - 10:00:02 PDT
