[gmx-users] Re: Problems with g_membed tools

[Mark Abraham]

On 11/11/11, LindaSong <wlsonglinda at gmail.com> wrote:

> Thank you for reply. I am sorry I didn't apply more information
> 
> There are only protein and lipids in my system now. I used the CHARMM ff. The topology is like this:
>     [ molecules ]
> ; Compound        #mols
> Protein_chain_A     1
> Protein_chain_A2    1
> POPC                1

 
I thought said you had 256 molecules of lipid... 
 

> 
> 
> In the grompp step, I used PME to calculate the electrostatic interaction, set the energy group to be Protein and the energy group exclusion to be 'Protein Protein'. I ignored the warning that grompp tells PME not calculating the long-distance interaction,because I think the long-distance interaction does not matter in this membrane insertion (Is this right?).

 
PME does calculate a long-range interaction. From your report, I don't think you are actually using PME, but it is probably irrelevant for membrane insertion. 
 

>  And then get the output tpr file. command line like this:
> 
>                 grompp -f membed.mdp -c complex_noLIG_m.pdb -p topol.top -o membed.tpr -maxwarn 1

 
Use of -maxwarn is a bad habit unless you really know what the problem is and the consequences of ignoring it... 
 

> 
> 
> Then I used the generated tpr file in the g_membed, command line like this:
> 
>                  g_membed -f membed.tpr -p topol.top -xyinit 0.1 -xyend 1.0 -nxy 1000
> 
> I selected the Protein to be embeded in and POPC the protein be embeded into;
> 
> Then the g_membed tells:
>       The estimated area of the protein in the membrane is 11.745 nm^2
> 
>       There are 1 lipids in the membrane part that overlaps the protein.
>       The area per lipid is inf nm^2.
>       Maximum number of lipids that will be removed is 0.

 
So there is something catastrophically wrong with the starting configuration...
 

> 
> 
> 
>       Will resize the protein by a factor of 0.100 in the xy plane and 1.000 in the z direction.
>      This resizing will be done with respect to the geometrical center of all protein atoms
>       that span the membrane region, i.e. z between 2.510 and 7.767
> 
>        Embedding piece 0 with center of geometry: 5.261230 5.247513 5.138600
> 
> 
>       -------------------------------------------------------
>     Program g_membed, VERSION 4.5.3
>     Source code file: gmx_membed.c, line: 688
> 
>        Fatal error:
>       Trying to remove more lipid molecules than there are in the membrane
>        For more information and tips for troubleshooting, please check the GROMACS
>        website at http://www.gromacs.org/Documentation/Errors
>       -------------------------------------------------------
> 
> So why the maximum number of lipids to be removed is 0? And I don't understand why in the topology file it says only 1 POPC? This is what I got when I used pdb2gmx to generate the topology file for POPC.
> 

 
It's hard to say without knowing what you did with pdb2gmx. You should be able to take your structure file with protein and POPC overlaid, pass that to pdb2gmx and get a correct .top. Depending on how the different POPC molecules are differentiated, you may need to choose pdb2gmx -chainsep differently.
 
Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20111111/50f66bbd/attachment.html>