[gmx-users] Covalent bonds between topology files

[wjallen]

Hello fellow gromacs users,

I am working with a protein that has an FAD cofactor that is covalently bound to
a specific residue. I am trying to model the protein in a style similar to Dr.
John E. Kerrigan's drug tutorial. I used pdb2gmx to create topology and
coordinate files for the protein, and the Dundee PRODRG server to create a
topology and coordinates (.gro) for the FAD molecule. I physically combined
the coordinate files, and I included the topology of the FAD molecule in the
header of the protein topology file:

; Include chain topologies
#include "protein.itp"
#include "FAD.itp"

and I included the FAD molecule at the end of the file:

[ molecules ]
; Compound #mols
Protein 1
FAD 1

Everything has been working well, the whole structure can be minimized and
processed so I know I have not made any errors yet. But, this is where I am
stuck. I need to covalently bind an atom in the FAD molecule to a residue in
the protein chain before I start minimizing anything. Any suggestions on how
to approach this would be greatly appreciated!

Cheers,
W. Allen

[Mark Abraham]

Post by wjallen
Everything has been working well, the whole structure can be minimized and
processed so I know I have not made any errors yet. But, this is where I am
stuck. I need to covalently bind an atom in the FAD molecule to a residue in
the protein chain before I start minimizing anything. Any suggestions on how
to approach this would be greatly appreciated!

All atoms involved in bonded interactions must be defined within the
same [ molecule ] segment. Thus you will need to go back and generate a
[ molecule ] segment that has this, and the atoms numbered correctly (so
you can't just concatenate).

Mark

[wjallen]

Thank you for your reply, Mark. That is what I was afraid of, that I would have
to physically combine the files. My only problem there is the format of the
topology file that comes from pdb2gmx is slightly different than that which
comes from the PRODRG server. For example, under the [ bonds ] section:

(from pdb2gmx):
...
5234 5236 2 gb_2
5237 5238 2 gb_4
...

(and from PRODRG):
...
5252 5254 1 0.140 334720.0 0.140 334720.0 CAY NBN
5254 5255 1 0.100 374468.0 0.100 374468.0 NBN HAF
...

I will try to decipher this and see what comes of it. Thankfully, the
renumbering of atoms was not a problem because of the reorder.pl script
available in the user contributions section. Again, thank you for your reply
Mark, I hope all that I said has made sense.

Cheers,
W. Allen

[Tsjerk Wassenaar]

Hi W. Allen,

This is not necessarily a problem. The gb_* labels can be found in the
ffG*bon.itp files, and are replaced by the values found in those files
upon preprocessing. If they both correspond to the same force field,
it's no problem to combine them with some parameters referred to by
codes and others written in full. You may have to worry though,
regarding the force field used for both files. This should be the same
for both.

Best,

Tsjerk

Post by wjallen
Thank you for your reply, Mark. That is what I was afraid of, that I would have
to physically combine the files. My only problem there is the format of the
topology file that comes from pdb2gmx is slightly different than that which
...
5234 5236 2 gb_2
5237 5238 2 gb_4
...
...
5252 5254 1 0.140 334720.0 0.140 334720.0 CAY NBN
5254 5255 1 0.100 374468.0 0.100 374468.0 NBN HAF
...
I will try to decipher this and see what comes of it. Thankfully, the
renumbering of atoms was not a problem because of the reorder.pl script
available in the user contributions section. Again, thank you for your reply
Mark, I hope all that I said has made sense.
Cheers,
W. Allen
_______________________________________________
gmx-users mailing list gmx-users at gromacs.org
http://www.gromacs.org/mailman/listinfo/gmx-users
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php

[Mark Abraham]

Post by wjallen
Thank you for your reply, Mark. That is what I was afraid of, that I would have
to physically combine the files. My only problem there is the format of the
topology file that comes from pdb2gmx is slightly different than that which
...
5234 5236 2 gb_2
5237 5238 2 gb_4
...
...
5252 5254 1 0.140 334720.0 0.140 334720.0 CAY NBN
5254 5255 1 0.100 374468.0 0.100 374468.0 NBN HAF
...

That's not a problem. The .itp files will have lines like "#define gb_2
<stuff>" which expands to something that resembles the latter examples,
except that the bonded function number is different so the parameter lists
need not match.

Mark