[gmx-users] Debugging 'inconsistent shifts'

[Kalev Takkis]

On 15 January 2015 at 19:44, Justin Lemkul <jalemkul at vt.edu> wrote:

>
>
> On 1/15/15 11:25 AM, Kalev Takkis wrote:
>
>> On 15 January 2015 at 00:09, Justin Lemkul <jalemkul at vt.edu> wrote:
>>
>>
>>>
>>> On 1/14/15 4:44 PM, Kalev Takkis wrote:
>>>
>>>  Hi, list!
>>>>
>>>> GROMACS complains about inconsistent shifts and suggests me to check my
>>>> topology. I have so done, but the problem is not immediately obvious.
>>>> How
>>>> can I make it tell me exactly which atoms/bonds it finds faulty?
>>>>
>>>>
>>>>  What were you doing when the error came up?  What was your exact
>>> command?
>>> What is your system?
>>>
>>>
>> Temperature equilibration, commands given were
>>       gmx grompp -f nvt.mdp -c simcell_minim.gro -p simcell.top -o
>> simcell_temp.tpr
>>       gmx mdrun -v -deffnm simcell_temp
>>
>> GROMACS version 5.0.4, ff OPLS-AA
>> Equilibration finishes without other problems, I can continue with the
>> next
>> step without the system blowing up. The error is not encountered during
>> minimisation, it finishes fine, energies are reasonable and no trace of
>> inconsistent shifts is found in the log.
>> System is chromatography column, silica slab with attached hydrocarbon
>> chains. The crystal surface is infinite in xz plane, so the mdp file
>> specifies
>> pbc             = xyz
>> periodic_molecules = yes
>>
>>
> Have you verified that all the bonded terms that span a periodic boundary
> are specified correctly?
>
>  The model is hacked together from silica unit cell and chains using
>> pdb2gmx, so it's entirely possible that there is indeed something wrong
>> with my system. I'd like to know where. When GROMACS encounters this error
>>
>
> The topology is the first and most likely source of error.
>
>  can I make it reveal more details? Which atoms/bonds are involved? Some
>> diagnostics tool or verbosity switch to mdrun maybe?
>>
>>
> How soon does the system blow up?  General troubleshooting advice is at
> http://www.gromacs.org/Documentation/Terminology/Blowing_Up#Diagnosing_an_
> Unstable_System
>

It doesn't. Noticed the message in the terminal window when I was trying to
improve the visual with trjconv after the production run. Then went back
through the logs and found that it was first reported in NVE equilibration.

>
> All GROMACS tools have a -debug flag that can be set, but you'll dump out
> a massive file with a ton of information you don't need.  If you're not
> used to reading through it, it can be more of an obstacle than a useful
> tool :)


No kidding. 1 step run -> 25MB log file. Fortunately grep was able to
extract the list of wonky atoms in a nice concise manner. Thank you for the
advice.


All the best,
Kalev

>
>
> -Justin
>
> --
> ==================================================
>
> 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.
>