[gmx-users] Re: Problems with CYS in FeS4 clusters

[David van der Spoel]

On Tue, 2002-12-10 at 10:58, Ernesto E. Di Iorio wrote:
> Hi there!
> Last week I sent an email to all gromacs users to find out how to treat an 
> FeS4 cluster. You kindly and promptly answered my email with instructions 
> on how to solve the problem. Accordingly, I have modified the ffG431a*itp 
> files to include the new bond, angle and dihedral types as well as the 
> proper charges for the CB and SG atoms of CYS1. Furthermore, I modifid the 
> specbond.dat file including a line like this
> CYS     SG      1       FE      FE      4       0.228   CYS1    FE
> When I try to generate a topology and the coordinates in gromacs format 
> under these conditions the four CYS residues involved in the formation of 
> the FeS4 cluster are missing the H atom attached to the backbone N. I have 
> tried to define a new type of CYS and also made an attempt using unmodified 
> ffG431a*itp files, but all I could generate is either fully protonated CYS 
> (namely with an H on both the N and the SG atoms) or CYS residues without 
> any H (neither on N nor on SG). Of course I can make the topology by hand 
You probably need another CYS type than CYS1, use CYS2, that has no
hydrogen on SG and should work straight in pdb2gmx


David