[gmx-users] Protein is having non-integral charge

[Mark Abraham]

On 7/04/2010 4:03 PM, sonali dhindwal wrote:
> Hello All
> I am using gromacs for simulating my protein I got after homology modelling.
> While doing this when I get toplogy file by using
> 0: GROMOS96 43a1 force field
> topology file I am getting has most of the residues having non-integral
> charge.
> Can someone help in this.
> As far as I searched for this, it is written that the charge should be
> integer.

Yes, each residue's charge needs to be an integer, and thus the whole 
charge also. Something's badly broken, either with what pdb2gmx is doing 
with your homology model, or how you're choosing termini. Look at the 
running counter of total charge on the right hand side of the listing in 
the [atoms] section of your .top file to see where things are going wrong.

Mark