[gmx-users] Large (600k particle) semi-isotropic lipid bilayer system transformed into vacuum-separated coordinates by grompp and mdrun

[George Pantelopulos]

Hi James,

Just now, I made the box about 6 nm smaller overall in every direction and
the system appears to be fitting nicely into the PBC now with no vacuums,
just as you thought it might.

I'll minimize an anneal this system -- I hope the answer is a simple and
almost alarmingly obvious as this.

Thanks!
George

On Thu, Sep 8, 2016 at 11:19 AM, James Graham <J.A.Graham at soton.ac.uk>
wrote:

> Hi George,
>
> That looks to me like the PBC box isn't quite tight enough around the
> starting configuration.  If you have a look at the box in vmd (pbc box) for
> the initial configuration, I suspect you'll see that it's off-centre with
> the corner of the box being right in the middle of the atoms, and also that
> it's big enough to contain all the atoms.  Your molecules seem to be in a
> 'pbc=whole' style, so you should expect some of them to overhang the edge
> of the box when it's properly centred.  I've had similar looking membranes
> when converting between GROMACS and LAMMPS.
>
> If that is indeed the issue, it can be fixed by making the box a little
> smaller in the input configuration using editconf.
>
> James
>
>
> On 08/09/16 16:07, George Pantelopulos wrote:
>
>> Hi Tsjerk,
>>
>> This system is from a MARTINI simulation that was mapped using the
>> CHARMM-GUI MARTINI to All-Atom Converter. The structure of the system is
>> completely transformed upon running in GMX, even running minimization for
>> one step.
>>
>> I'll post a few quick screencaps of the system as it was input, after one
>> step of minimization, and after annealing here:
>> https://s18.postimg.org/n0jljxow9/600000particle_splitting.png
>>
>> Thank you!
>> George
>>
>> On Thu, Sep 8, 2016 at 10:11 AM, Tsjerk Wassenaar <tsjerkw at gmail.com>
>> wrote:
>>
>> Hi George,
>>>
>>> Oh, quadrants over PBC, so it just completely ruptured. It means the area
>>> per lipid is really off in your starting structure. Where did you get it
>>> from?
>>>
>>> Cheers,
>>>
>>> Tsjerk
>>>
>>> On Sep 8, 2016 3:54 PM, "George Pantelopulos" <gpantelopulos at gmail.com>
>>> wrote:
>>>
>>> Dear Tsjerk and Erik,
>>>>
>>>> Thank you for the responses!
>>>>
>>>> The initial structure is a complete bilayer with no vacuum in a
>>>>
>>> rectangular
>>>
>>>> box.
>>>>
>>>> The quadrants are due to distribution over cores/nodes. You can try EM
>>>>
>>> on a
>>>
>>>> single core for a few steps, but it seems that the system is too
>>>>>
>>>> stretched
>>>>
>>>>> anyway, so I think that won't really solve your problem.
>>>>>
>>>>
>>>> I had thought this as well, at first, but even running on a single
>>>> processor for one step produces this problem -- it is for this reason
>>>>
>>> that
>>>
>>>> I think it has something to do with how GROMACS is interpreting the
>>>>
>>> system.
>>>
>>>> Is this not just a visualisation issue because of periodic boundary
>>>>
>>>>> conditions? Are you sure that the vacuum was not present in the input
>>>>> coordinate file?
>>>>>
>>>>
>>>> I am sure that there was no vacuum present in the input coordinate file.
>>>>
>>> I
>>>
>>>> had thought that it might be a bizarre visualization issue, but after a
>>>> very rigorous minimization and annealing from 0K, the system not only
>>>> remains split into quadrants, but the edges of each quadrant deform from
>>>> squares into more smoothly curved shapes. If I look at the system in VMD
>>>> with periodic images, it seems like the outer edges of all of the
>>>>
>>> quadrants
>>>
>>>> align perfectly to make one complete bilayer with the nighboring over
>>>> the
>>>> neighboring, but the vacuums are still there.
>>>>
>>>> On Thu, Sep 8, 2016 at 5:42 AM, Erik Marklund <erik.marklund at kemi.uu.se
>>>> >
>>>> wrote:
>>>>
>>>> Dear George,
>>>>>
>>>>> Is this not just a visualisation issue because of periodic boundary
>>>>> conditions? Are you sure that the vacuum was not present in the input
>>>>> coordinate file?
>>>>>
>>>>> Kind regards,
>>>>> Erik
>>>>>
>>>>> On 7 Sep 2016, at 23:58, George Pantelopulos <
>>>>>>
>>>>> gpantelopulos at gmail.com>
>>>
>>>> wrote:
>>>>>
>>>>>> Dear all,
>>>>>>
>>>>>> I have been trying to equilibrate a very large lipid bilayer system
>>>>>>
>>>>> for
>>>
>>>> the
>>>>>
>>>>>> past few days, but I have not been able to move past a very curious
>>>>>>
>>>>> error
>>>>
>>>>> that happens when producing a tpr file and running miminization or
>>>>>>
>>>>> MD.
>>>
>>>> The system ends up being split in to 8 quadrants, leaving sizeable
>>>>>>
>>>>> vacuums
>>>>>
>>>>>> between each quadrant, and the coordinates of the system are
>>>>>>
>>>>> transposed.
>>>>
>>>>> I've tried to resolve this via various combinations of minimization
>>>>>>
>>>>> and
>>>
>>>> equilibration schemes but have had no success. At this point, I
>>>>>>
>>>>> think I
>>>
>>>> may
>>>>>
>>>>>> only turn to the mailing list for advice.
>>>>>>
>>>>>> Does anyone have an idea of what might be going on and how I could
>>>>>>
>>>>> resolve
>>>>>
>>>>>> this?
>>>>>> Thank you for the help,
>>>>>> George Pantelopulos
>>>>>> --
>>>>>> 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.
>>>>>
>>>>> --
>>>>> 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.
>>>>>
>>>>> --
>>>> 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.
>>>>
>>>> --
>>> 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.
>>>
>>>
> --
> James Graham - PhD Student
> Institute for Complex Systems Simulation (ICSS)
> Computational Systems Chemistry
> University of Southampton, UK
>
>
> --
> 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.
>