[gmx-users] pdb2gmx with force field option

[Mark Abraham]

> As my first Gromacs simulation I am trying to simulate a protein with K+
> ion.
> When I use pdb2gmx with GROMACS96 43a2, it complains about the K+ ion and
> does not generate the topology file.
>
> When I remove the K+ ion, I get a topology file, but I need the K+ for the
> simulation.  Can I add the K+ ions at the end of the .gro files and
> proceed?

No that won't work. You need to choose a force field that has been
parameterized to include your protein and your ion. I don't know if this
one has.

> 1. How can I run pdb2gmx and tell it to look for the force filed
> information
> for K in a different file?  I tried -ff but got unclear message?

You can use the #include mechanism in your .top if you have an .itp file
for potassium.

> 2.  When I searched in the share/top directrory I found that each force
> field name with several extensions:.itp, .dhb and .atp.  I only have a
> K.itpfile, is that enough?

Yes, if it's complete. Have a read of chapter five of the manual.

> 3.  To insert the protein into a lipid I used genbox command:
>      genbox -cp protein.pdb -cs lipid.pdb
>
> The protein sinks deep into the lipid and I need to shift the protein back
> up? What is the best to do it without overlapping with the solvent
> molecules?

Start with MD on a simple system while you work out how things work. First
just your protein. Then with the ion. Then worry about the lipid.

> 4. What is a better way of doing step 3?

I don't know.

Mark