[gmx-users] annealing

[Nuno R. L. Ferreira]

Luciano wrote:
>  Hi gromacs users,
> 
>  I'm with problem:
> 
>  I have a protein file generated by pdb2gmx, where the file in *.pdb
> does not orginate from an X-ray structure. It was generated by Sybyl
> program.

Did pdb2gmx report any error about unknown atoms or residues? If not, 
all OK till this point.

> 
>  Well, I have the following mdp file:
> 
> cpp                     = /lib/cpp
> dt                      = 0.001  ; ps
> nsteps                  = 100000
> nstcomm                 =  1                        ; reset c.o.m. motion
> nstlog                  = 10
> nstlog                  = 100
> nstenergy               = 100
> energygrps              = system
> ns_type                 = grid
> coulombtype             = PME
> fourierspacing          = 0.12
> annealing               = yes
> tinit                   = 0
> ref_t                   = 1000    ; K
> zero_temp_time          = 50

I think 50 ps it's a short time for SA. Check recent similar questions 
posed on gmx e-mail list.
If you want a real minimum, you have to do this slowly IMHO. You are

> tcoupl                  = Berendsen
> tc-grps                 = system
> tau_t                   = 0.001
> pcoupl                  = Berendsen
> tau_p                   = 0.05
> compressibility         = 4.5e-5 ; ???

Isothermal Compressibility of your solvent. Default value is for water. 
Usually you don't change this.

> ref_p                   = 1.0    ; atm

Actually this comes in bar (1atm aprox. 1.01 bar). Just being "picuinhas".

> gen_vel                 = no

Do you have velocities at 1000k in your input structure file/trajectory?

> constraints             = none
> rlist                   = 1.0
> rcoulomb                = 1.0
> rvdw                    = 1.0
> 
>  The grompp program generate the tpr file and when I try mdrun the
> program shows:
> 
>  Getting Loaded...
> Reading file octa20orig.tpr, VERSION 3.1.4 (single precision)
> Reading file octa20orig.tpr, VERSION 3.1.4 (single precision)
> Loaded with Money
> 
> starting mdrun 'w-gliadin'
> 100000 steps,    100.0 ps.
> 
> step 0Segmentation fault
> 
>  I think that was the posre.itp generated by pdb2gmx, but not have
> certainty. Why the segmentation fault?!

I think not, because you don't have any define entry in your mdp file.
Please be more precise.

Um olá aos meus companheiros Lusos.
Tchau,
Nuno

> 
>  Cheers,
> 
>  Luciano 
>  
> 
> ######################################################
>                Luciano Tavares da Costa
>              Mestre em Engenharia Química
>    icq #232852036 ; ltcosta at universiabrasil.net  
>     Doutorando Laboratório de Biofisicoquímica
>          Instituto de Química - USP - Brazil
>              
> ######################################################
> 
> 
> ------------------------------------------------------------------------
> Universia, o portal dos universitários.
> Serviços, informações e conteúdo acadêmico na maior rede 
> universitária do mundo. (www.universiabrasil.net)
>