[gmx-users] Error running simulations with the protein frozen

Mon Mar 28 02:44:28 CEST 2016

Hi Justin,

Thanks for the reply! I have passed the minimization and NVT equilibration
by relaxing the protein for just a little bit then freezing it. But, I had
a realization that the freeze algorithm is not compatible with pressure
scaling in NPT... so I can't do my free energy calculation in NPT ensemble.
I think I have to switch to restraints after all.

Thanks again,
Nuo

On Sun, Mar 27, 2016 at 5:36 PM, Justin Lemkul <jalemkul at vt.edu> wrote:

>
>
> On 3/26/16 1:27 AM, Nuo Wang wrote:
>
>> Hi Justin,
>>
>> This "freeze-protocol" is published here in Methods:
>> http://pubs.acs.org/doi/full/10.1021/acs.jctc.5b00684
>> My goal is to use the MD-calculated free energy to validate the
>> electrostatic solvation free energies produced by software like APBS. In
>> APBS, the protein takes one conformation (typically a PDB structure) and
>> the solvent is modeled by mean-field theory. The only rigorous way to
>> compare MD to the "single-conformation" APBS result is by freezing the
>> protein and sampling only the solvent.
>>
>> The system does look artificial but from an algorithmic point of view,
>> my protocol should be very doable. The protein is frozen, so its high
>> internal energy doesn't matter. And water is reasonably placed around the
>> protein by Gromacs, water themselves should pass the minimization safely.
>>
>> To give you more information, I have done the exactly same thing in NAMD
>> --
>> freezing the crystal structure protein -- and it works just fine. But now
>> I
>> have to switch to Gromacs to use the free energy calculation procedure I
>> need. (Long time NAMD user, first time Gromacs user)
>>
>> Last, I have already tried to use restraints, simulation runs fine, but it
>> is not what I want to do.
>>
>> I think this is a input file setup issue, not an issue with the physics of
>> the seemingly artificial system.
>>
>>
> There is nothing wrong with the input file; minimization scripts are very
> simple.  Of course, your system will never converge to Fmax < 10 with
> single precision and only 100 steps, but that's not really the issue here.
> You may need to look into using energrp_excl when freezing coordinates,
> because there are artificial contributions to the forces when freezing.
>
> -Justin
>
>
> Thanks ahead for any further help on this!
>>
>> Thanks,
>> Nuo
>>
>> On Fri, Mar 25, 2016 at 8:32 PM, Justin Lemkul <jalemkul at vt.edu> wrote:
>>
>>
>>>
>>> On 3/25/16 9:42 PM, Nuo Wang wrote:
>>>
>>> Hi Gromacs admins,
>>>>
>>>> I am trying to run MD simulations of protein solvated in water. I want
>>>> to
>>>> completely freeze the protein at its crystal conformation and only
>>>> sample
>>>> the solvent degrees of freedom (for comparison with Poisson-Boltzmann
>>>> calculations that only use the crystal structure). I tried to follow the
>>>> typical minimization-equilibration-production procedure described here:
>>>>
>>>>
>>>> http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/lysozyme/index.html
>>>>
>>>> But I get this error at the energy minimization step:
>>>> "Fatal error:
>>>> step 11: Water molecule starting at atom 60322 can not be settled.
>>>> Check for bad contacts and/or reduce the timestep if appropriate."
>>>>
>>>> If I take away the freeze protein commands in my minimize.mdp, then no
>>>> error occurs:
>>>> freezegrps    = Protein
>>>> freezedim     = Y Y Y
>>>>
>>>> I checked my atom 60322, it is next to the protein, but it is neither
>>>> clashing with anyone nor moving rapidly in my trajectory. Reducing
>>>> timestep
>>>> by 1000 times won't help either, error just occurs at a later step.
>>>>
>>>> I wonder how can I resolve this problem?
>>>>
>>>>
>>> Freezing is a severe perturbation on the system, and taking a protein out
>>> of its crystal environment, dunking it in water, and expecting that
>>> conformation to be happy (or even sensible) is a bad assumption.  If you
>>> want to preserve the overall conformation, try something gentler like
>>> normal position restraints.
>>>
>>> -Justin
>>>
>>>
>>> The script that I used is shown here, the input files in the script are
>>>
>>>> linked below:
>>>> ### Generate topology
>>>> echo 9  | gmx pdb2gmx -f complex.pdb -o complex.gro -p complex.top
>>>> -merge
>>>> all -ignh -water tip3p
>>>> ### Generate box information
>>>>             gmx editconf -f complex.gro -o complex_newbox.gro -bt cubic
>>>> -c
>>>> -box 11 11 11
>>>> ### Add water
>>>>             gmx solvate -cp complex_newbox.gro -cs spc216.gro -o
>>>> complex_solvated.gro -p complex.top
>>>> ### Prepare ion parameters
>>>>             gmx grompp -f ions.mdp -c complex_solvated.gro -p
>>>> complex.top
>>>> -o
>>>> ions.tpr
>>>> ### Add ions
>>>> echo 13 | gmx genion -s ions.tpr -o complex_solvated_ions.gro -p
>>>> complex.top -pname NA -nname CL -neutral -conc 0
>>>> ### Generate index file
>>>> echo q  | gmx make_ndx -f complex_solvated_ions.gro -o index.ndx
>>>> ### Preparation for energy minimization for the non-protein atoms
>>>>             gmx grompp -f minimize.mdp -c complex_solvated_ions.gro -n
>>>> index.ndx -p complex.top -o em.tpr
>>>> ### Energy minimization for the non-protein atoms
>>>>             gmx mdrun -v -deffnm em -nt 1
>>>>
>>>> complex.pdb:
>>>>
>>>>
>>>> https://drive.google.com/file/d/0B8oskrjCa2EMQURKd0J5bXdNRk0/view?usp=sharing
>>>> ions.mdp:
>>>>
>>>>
>>>> https://drive.google.com/file/d/0B8oskrjCa2EMS3M4cTdkRXRTRE0/view?usp=sharing
>>>> minimize.mdp:
>>>>
>>>>
>>>> https://drive.google.com/file/d/0B8oskrjCa2EMU0hZT0lxXzZuM1U/view?usp=sharing
>>>>
>>>> Thanks a lot!
>>>> Nuo
>>>>
>>>>
>>>> --
>>> ==================================================
>>>
>>> Justin A. Lemkul, Ph.D.
>>> Ruth L. Kirschstein NRSA Postdoctoral Fellow
>>>
>>> Department of Pharmaceutical Sciences
>>> School of Pharmacy
>>> Health Sciences Facility II, Room 629
>>> University of Maryland, Baltimore
>>> 20 Penn St.
>>> Baltimore, MD 21201
>>>
>>> jalemkul at outerbanks.umaryland.edu | (410) 706-7441
>>> http://mackerell.umaryland.edu/~jalemkul
>>>
>>> ==================================================
>>> --
>>> Gromacs Users mailing list
>>>
>>> * Please search the archive at
>>> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
>>> posting!
>>>
>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>>
>>> * For (un)subscribe requests visit
>>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>>> send a mail to gmx-users-request at gromacs.org.
>>>
>>>
> --
> ==================================================
>
> Justin A. Lemkul, Ph.D.
> Ruth L. Kirschstein NRSA Postdoctoral Fellow
>
> Department of Pharmaceutical Sciences
> School of Pharmacy
> Health Sciences Facility II, Room 629
> University of Maryland, Baltimore
> 20 Penn St.
> Baltimore, MD 21201
>
> jalemkul at outerbanks.umaryland.edu | (410) 706-7441
> http://mackerell.umaryland.edu/~jalemkul
>
> ==================================================
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>