[gmx-users] Distance Restraints: Inconsistent Shifts, LINCS Warnings, Number of grid cells is zero, XTC error

Mark Abraham Mark.Abraham at anu.edu.au
Sat Oct 10 00:54:05 CEST 2009


João M. Damas wrote:
> What do you mean by mutually inconsistent? From what I understand, they 
> aren't.

I mean a set for which there is no molecular arrangement that satisfies.

> Also, [bonds] type 6 won't work for what I want, since I don't want to 
> keep the molecules at a fixed distance but to prevent them from coming 
> closer, i.e. I want a potential that depends on the distance between the 
> atoms.
> 
> Testing the distance restraints can be quite exhaustive when you have 
> about 160 restraints, and I am considering just the test case I'm 
> trying, not the actual system which will have some more organic 
> molecules, which will lead to a greater amount of combination of 
> restraints...

#ifdef is your friend here...

[ distance_restraints ]

<set to test now >

#ifdef BLAHBLAH

<set to test later>

#endif

Now you just move the #ifdef to rearrange your test cases.

I might start with an even smaller test case until you prove you can 
deal with both the input formatting issues and the physics issues with 
it. Testing will be faster, for starters.

Mark

> On Fri, Oct 9, 2009 at 12:51 AM, Mark Abraham <Mark.Abraham at anu.edu.au 
> <mailto:Mark.Abraham at anu.edu.au>> wrote:
> 
>     João M. Damas wrote:
> 
>         Gromacs users,
> 
>         I have a small test system of about 2500 atoms which consists on
>         a few organic molecules in a box of water. I'm trying to use
>         distance restraints to keep each molecule apart from each other
>         and I've failed until now.
> 
>         I'll start by saying that the simulation runs smoothly without
>         using distance restraints and that I've created a single
>         molecule topology that describes all organic molecules, allowing
>         the use of distance restraints. By applying just one distance
>         restraint between two atoms of different molecules, the
>         simulation still runs without problems. When I apply all the
>         distance restraints I want, the system crashes in the first few
>         steps with an XTC error, which is preceded by a few inconsistent
>         shifts and some LINCS warnings. Varying the number of restraints
>         that I use, makes the crash happen at later steps (500ish),
>         meanwhile giving a lot of inconsistent shifts before crashing
>         with a fatal error: Number of grid cells is zero. I know it
>         seems I'm experimenting blindly, but varying the number of
>         restraints was the first that occurred to me.
>         I've searched the list for similar problems and tried a few more
>         tests. I've turned off LINCS and reduced the time step and it
>         didn't solve the problem, the box blew up after a few steps due
>         to pressure problems. Since it could be a PBC problem, I tried
>         to increase the size of the box, but it didn't solve the
>         problem. Turning of PBC did solve the problem, but I really need
>         PBC. I've also changed the low, up1 and up2 to values inferior
>         to half the box size (I don't even care about the up1 and up2
>         values in this application, and I was setting them to 5.0 nm
>         before), but it didn't solve the problem either. I'm a little on
>         a dead end here and I was hoping anyone could give me some insight.
> 
>         By the way, I'm using GROMACS 4.
> 
> 
>     The simplest explanation is that your sets of distance restraints
>     are mutually inconsistent. This destabilizes the system and
>     eventually leads to your symptoms. Consider using [bonds] type 6,
>     per advice of manual section 4.3.4. Either way, if you add these
>     restraints one by one after testing that the previous set seemed to
>     work OK (didn't crash, trajectory looks right), you may uncover your
>     issue.
> 
>     Mark
>     _______________________________________________
>     gmx-users mailing list    gmx-users at gromacs.org
>     <mailto:gmx-users at gromacs.org>
>     http://lists.gromacs.org/mailman/listinfo/gmx-users
>     Please search the archive at http://www.gromacs.org/search before
>     posting!
>     Please don't post (un)subscribe requests to the list. Use the www
>     interface or send it to gmx-users-request at gromacs.org
>     <mailto:gmx-users-request at gromacs.org>.
>     Can't post? Read http://www.gromacs.org/mailing_lists/users.php
> 
> 
> 
> 
> -- 
> João M. Damas
> PhD Student
> Protein Modelling Group
> ITQB-UNL, Oeiras, Portugal
> Tel:+351-214469613
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the 
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php



More information about the gromacs.org_gmx-users mailing list