[gmx-users] Re: Problems prior to running lipid-protein simulation

[Bruce Milne]

gmx-users-request at gromacs.org wrote:
> Bruce Milne wrote:
>   
>> > You don't have to run pdb2gmx for the whole system at any point 
>>     
> Probably true.
>> > (it's 
>> > not going to work anyway unless you add the POPC parameters to all the 
>> > force field files and then add the POPC residue to the  .rtp file so 
>> > that  pdb2gmx can recognize it).
>>     
> Not quite true. The contents of the .itp files don't matter to pdb2gmx 
> since it writes a .top file that #includes them. So you need to have a 
> working POPC entry in an .rtp file for pdb2gmx, and later on, when 
> using grompp, you'll need the former contents of the .rtp file to 
> correspond to stuff the .itp files. The latter you need even without 
> using pdb2gmx, of course.

Good point. I was only considering the case where you want pdb2gmx to 
provide you with a complete .top file that needs no additions and takes 
all its parameters from the same place (as you would get for a protein 
for example).

Anyway, if you have the .itp files then I think it is easier to simply 
add the required #include lines to the protein.top file so that it 
covers all the molecule types in your coordinate file without trying to 
get pdb2gmx to recognize the lipids.

Cheers,

Bruce

-- 
"Hamewith - the road that's never dreary"

Dr. Bruce F. Milne
CEQOFFUP
Faculdade de Farmácia
Universidade do Porto
Rua Aníbal Cunha - 164
4050-047
Porto
Portugal

email: bfmilne at ff.up.pt