[gmx-users] Use of restraints gives segmentation fault

[Justin A. Lemkul]

Anna Duncan wrote:
> 
> I'm trying to energy minimise a protein structure.  In one area of the 
> structure there are some salt bridges and when I do an energy 
> minimisation without any restraints these are disrupted, which is 
> something I want to avoid.  To that end, I've tried restraining the 
> residues that are involved, first with position restraints and then with 
> distance restraints, but in both cases I get a segmentation fault in the 
> same place.
> The command I issue looks like:
> grompp -f em.mdp -c 1okc_ex.cg.gro -p 1okc_ex.top -o 1okc_min_sbRes -v 
> -n salt_bridges.ndx
> 
> and the output:
> Ignoring obsolete mdp entry 'title'
> Ignoring obsolete mdp entry 'cpp'
> Replacing old mdp entry 'unconstrained_start' by 'continuation'
> 
> Back Off! I just backed up mdout.mdp to ./#mdout.mdp.10#
> checking input for internal consistency...
> 
> NOTE 1 [file em.mdp, line unknown]:
>   For energy conservation with switch/shift potentials, rlist should be 0.1
>   to 0.3 nm larger than rcoulomb.
> 
> 
> NOTE 2 [file em.mdp, line unknown]:
>   For energy conservation with switch/shift potentials, rlist should be 0.1
>   to 0.3 nm larger than rvdw.
> 
> processing topology...
> Generated 0 of the 465 non-bonded parameter combinations
> Segmentation fault
> 
> --------------------------------------------------------------------------------------
> 
> Like I say, when I do this without the restraints, grompp and mdrun run 
> fine.  The only differences, then, between the input is the line in em.mdp:
> define = -DDISTRESB
> 
> The topoolgy file, 1okc_ex.top looks like:
> #include "martini_v2.1.itp"
> #include "1okc_ex.itp"
> #ifdef POSRESB
> #include posre_saltbridges.itp
> #endif
> #ifdef DISTRESB
> #include distre_sb.itp
> #endif
> 

You have the wrong syntax.  You need " " around your .itp files in the #include 
statements.  Note how your first two #includes look different from your last two :)

-Justin

> [ system ]
> 1OKC_EX SIMULATION
> 
> [ molecules ]
> Protein 1
> 
> --------------------------------------------------------------------------------------
> 
> The position restraints file looks like: 
> ; position restraints for r_28_r_31_r_133_r_136_r_230_r_233 of God Rules 
> Over Mankind, Animals, Cosmos and Such
> 
> [ position_restraints ]
> ;  i funct       fcx        fcy        fcz
>   52    1       1000       1000       1000
>   53    1       1000       1000       1000
>   59    1       1000       1000       1000
>   60    1       1000       1000       1000
>   61    1       1000       1000       1000
>  296    1       1000       1000       1000
>  297    1       1000       1000       1000
>  303    1       1000       1000       1000
>  304    1       1000       1000       1000
>  305    1       1000       1000       1000
>  505    1       1000       1000       1000
>  506    1       1000       1000       1000
>  511    1       1000       1000       1000
>  512    1       1000       1000       1000
>  513    1       1000       1000       1000
> 
> --------------------------------------------------------------------------------------
> 
> and the distance restraints file looks like:
> ; distance restraints for h1-h3_sb of God Rules Over Mankind, Animals, 
> Cosmos and Such
> 
> [ distance_restraints ]
> ;   i     j ? label      funct         lo        up1        up2     weight
>    53   395 1     0          1    2.00181    2.20181    3.20181          1
>   297   513 1     0          1   0.462702   0.662702     1.6627          1
>    61   506 1     0          1     0.3818     0.5818     1.5818          1
> 
> --------------------------------------------------------------------------------------
> 
> 
> I can't work out what is giving the segmentation fault because I think 
> the restraint files are in the right format and I can't think of what 
> else would go wrong at this stage for a segmentation fault to occur.
> 
> In the grompp run that works (without any restraints) the line after 
> where the segmentation error occurs in the other cases looks like:
> Excluding 1 bonded neighbours molecule type 'Protein'
> 
> Many thanks,
> Anna
> 

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================