[gmx-users] amb2gmx for big systems

[Florian Haberl]

Hi,

On Thursday, 31. July 2008, Rebeca García Fandiño wrote:
> Hello,
> I am trying to convert a membrane-protein system from amber to gromacs
> using the script amb2gmx.pl
> (http://chemistry.csulb.edu/ffamber/tools.html). It is the dry system, only
> the protein and a membrane composed by DOPC, whose parameters I have got
> from Gaff.
>
> When I try:
>
> ./amb2gmx.pl --prmtop dry_system.top --crd dry_system.crd --outname
> dry_system_gmx
>
> I get an error related to the “*” symbols in the pdb (I have 253.000
> atoms): (…)
> Argument " NATOM or " isn't numeric in multiplication (*) at ./amb2gmx.pl
> line 200, <PDB> line 1. Argument "NRES is to" isn't numeric in
> multiplication (*) at ./amb2gmx.pl line 201, <PDB> line 1. Argument "o
> big:" isn't numeric in multiplication (*) at ./amb2gmx.pl line 202, <PDB>
> line 1. substr outside of string at ./amb2gmx.pl line 205, <PDB> line 1.
> Use of uninitialized value in addition (+) at ./amb2gmx.pl line 205, <PDB>
> line 1. substr outside of string at ./amb2gmx.pl line 208, <PDB> line 1.
> Use of uninitialized value in addition (+) at ./amb2gmx.pl line 208, <PDB>
> line 1. substr outside of string at ./amb2gmx.pl line 209, <PDB> line 1.
> Use of uninitialized value in addition (+) at ./amb2gmx.pl line 209, <PDB>
> line 1. substr outside of string at ./amb2gmx.pl line 212, <PDB> line 1.
>
>
> Is there a way to transform amber systems to gromacs in spite of the “*”
> symbols? Has anyone transformed this type of systems? How could I save the
> problem?

I`ll tried systems with 45k atoms and it works, i don`t know if the script 
works with more than 100k atoms depends on the numbering of your pdb file. 
The pdb file format is a rather old fixed format (FORTRAN like) and counts 
only up to 99999 atoms in this fixed format. In my amb2gmx.pl version 
(amb2gmx.pl v.001 2006-01-14) all things like resname, atomname and so on are 
read by fixed length so you have to modify this fixed length or split the pdb 
file.

Another possibility is to use the amberff implementation for gromacs setup the 
protein and membrane with gromacs itself. The generated topology of your 
membrane with gaff can easily be transfered to gromacs itp/ gro file.



>
> Thank you very much for your help in advance,
>
> Rebeca García Fandiño
> Parc Cientific de Barcelona
>
>
>
> _________________________________________________________________
> Herramientas para combatir la crisis. MSN Dinero
> http://dinero.es.msn.com/


Greetings,

Florian


-- 
-------------------------------------------------------------------------------
 Florian Haberl                        
 Computer-Chemie-Centrum   
 Universitaet Erlangen/ Nuernberg
 Naegelsbachstr 25
 D-91052 Erlangen
 Telephone:  	+49(0) − 9131 − 85 26573
 Mailto: florian.haberl AT chemie.uni-erlangen.de
-------------------------------------------------------------------------------