[gmx-users] gmx wham (again)

Alex nedomacho at gmail.com
Thu Apr 19 00:30:12 CEST 2018


An otherwise perfectly well-behaved setup is doing something catastrophic.
The pull code is:

; Pull code
pull                 = yes
pull-coord1-type     = umbrella
pull_ngroups            = 2
pull_ncoords            = 1
pull_group1_name        = K
pull_group2_name        = CNT
pull_coord1_geometry    = cylinder
pull-cylinder-r         = 1.0
pull-group2-pbcatom     = 1343
pull_coord1_groups      = 1 2
pull-coord1-vec         = 0 0 1
pull_coord1_rate        = 0.0
pull_coord1_k           = 1000          ; kJ mol^-1 nm^-2
pull_coord1_start       = yes           ; define initial COM distance > 0

Segfaults without error messages within 20-50 ps. I suspected that
pull-coord1-vec is somehow hating zero coord1_rate, so I changed that to
1.0e-9.

The result is:

***
Program:     gmx mdrun, version 2018
Source file: src/gromacs/mdlib/sim_util.cpp (line 776)
MPI rank:    0 (out of 4)

Fatal error:
Step 40700: The total potential energy is -nan, which is not finite. The LJ
and electrostatic contributions to the energy are 15121 and -108386,
respectively. A non-finite potential energy can be caused by overlapping
interactions in bonded interactions or very large or Nan coordinate values.
Usually this is caused by a badly- or non-equilibrated initial
configuration,
incorrect interactions or parameters in the topology.
***

No indication of exploding system at step 40000, all energies looking legit.

Any thoughts?

Alex




On Wed, Apr 18, 2018 at 4:00 PM, Alex <nedomacho at gmail.com> wrote:

> ... and everything is segfaulting, though worked perfectly fine until the
> 'cylinder' thing.
>
> On Wed, Apr 18, 2018 at 3:45 PM, Alex <nedomacho at gmail.com> wrote:
>
>> Which then would be = 0 0 1, correct? This though will have to coexist
>> with pull_coord1_rate        = 0.0
>>
>> Is that reasonable?
>>
>> Thanks,
>>
>> Alex
>>
>> On Wed, Apr 18, 2018 at 2:11 PM, Justin Lemkul <jalemkul at vt.edu> wrote:
>>
>>>
>>>
>>> On 4/18/18 4:08 PM, Alex wrote:
>>>
>>>> In addition, we seem to have a different problem:
>>>>
>>>
>>> Cylinder geometry requires the use of pull-coord1-vec, not
>>> pull-coord1-dim.
>>>
>>> -Justin
>>>
>>>
>>> -------------------------------------------------------
>>>> Program:     gmx grompp, version 2018
>>>> Source file: src/gromacs/gmxpreprocess/readpull.cpp (line 234)
>>>>
>>>> Fatal error:
>>>> With pull geometry cylinder the pull vector can not be 0,0,0
>>>>
>>>>
>>>> On Wed, Apr 18, 2018 at 6:34 AM, Justin Lemkul <jalemkul at vt.edu> wrote:
>>>>
>>>>
>>>>> On 4/18/18 3:17 AM, Alex wrote:
>>>>>
>>>>> I suppose this question is mostly for Justin...
>>>>>>
>>>>>> Let me remind what I am dealing with and ask if my idea is correct.
>>>>>>
>>>>>> I have a rectangular membrane in XY with a pore at (X/2, Y/2) in water
>>>>>> and want to get the Gibbs free energy curve for an ion. For this, I
>>>>>> have a
>>>>>> bunch of starting configurations at (X/2, Y/2) and Z varying between
>>>>>> some
>>>>>> -z0 and z0. The bias in the "fake" pull mdp is applied as N N Y. Near
>>>>>> the
>>>>>> membrane, this means the entire plane is sampled, which adds
>>>>>> contributions
>>>>>> I am not interested in. I want the pore and a small region of the
>>>>>> membrane
>>>>>> around it, not the membrane, given its properties.
>>>>>>
>>>>>> So, I applied weak (k=50) in-plane restraint to the ion for each of
>>>>>> the
>>>>>> sampled configurations -- to keep the sampling region a bit closer to
>>>>>> the
>>>>>> pore. The results look completely different, but they finally make
>>>>>> very
>>>>>> good qualitative sense. The ion still walks around within a small
>>>>>> disk, but
>>>>>> not much -- this tentatively makes me happy.
>>>>>>
>>>>>> Would you believe the results obtained this way, assuming otherwise
>>>>>> proper setup?
>>>>>>
>>>>>>
>>>>>> This is the exact situation that the "cylinder" pull geometry is
>>>>> designed
>>>>> to model. I suggest going that route rather than ad hoc restraints,
>>>>> which
>>>>> will influence the PMF and should be otherwise accounted for.
>>>>>
>>>>> -Justin
>>>>>
>>>>> --
>>>>> ==================================================
>>>>>
>>>>> Justin A. Lemkul, Ph.D.
>>>>> Assistant Professor
>>>>> Virginia Tech Department of Biochemistry
>>>>>
>>>>> 303 Engel Hall
>>>>> 340 West Campus Dr.
>>>>> Blacksburg, VA 24061
>>>>>
>>>>> jalemkul at vt.edu | (540) 231-3129
>>>>> http://www.thelemkullab.com
>>>>>
>>>>> ==================================================
>>>>>
>>>>> --
>>>>> 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.
>>> Assistant Professor
>>> Virginia Tech Department of Biochemistry
>>>
>>> 303 Engel Hall
>>> 340 West Campus Dr.
>>> Blacksburg, VA 24061
>>>
>>> jalemkul at vt.edu | (540) 231-3129
>>> http://www.thelemkullab.com
>>>
>>> ==================================================
>>>
>>> --
>>> 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.
>>>
>>
>>
>


More information about the gromacs.org_gmx-users mailing list