CHARMM Development Project Forums User Discussion & Questions Setup, I/O, and Basic questions patch necessary?

Hello,

How important is to use "S-S" patch in a simulation e.g. in a structure which is built up from homology model and x-ray structure is not known. I would think that S-S bond is so strong, patch is not necessary to be used to define a covalent bond. Is this true?

thanks, preeti

The Cys residue in CHARMM is defined as the -SH form by default. Cys disulfide bridges are only created via the explicit use of the DISU patch, which removes the H atoms and creates the important inter-chain S-S bond. It is a mistake to omit the patch for a protein with known disulfide bridges.

Rick, Thanks for clarifying that for me. preeti

Hi Rick, I found this thread which is of immediate interest to me, so I hope you and preeti do not mind if I barge in with two questions of my own.

Some background
I am working with a coiled-coil parallel dimer. I am generating my .psf .crd and .pdb files for the dimer by GENErating each chain separately from a pdb xtal structure and appending the two chains before minimization. The published structure shows a disulfide bond, although at the published resolution (2.3angstroms), the author admits that it is not completely clear that the structure is cross-linked. This molecule is known to be experimentally oxidized though it is not known if the cross-link is functionally relevant.

First question:
After generating, appending and minimizing my structure for MD, the charmm generated structure that is produced is cross-linked (ie. -S-S- bond is present) like the original pdb structure file. Given your comment here, do I still need to include a PATCh DISU ... command when appending the chains eventhough the generated structure is actually disulfide cross-linked?

Second question:
In order to remove the cross-link, I am assuming that I have to manually rotate about the C-S bond and re-minimimize the structure. If there a simpler way to do this, could you point me to it?

How do you know it is cross-linked? Have you verified that the S-S bond exists in the PSF? If you're using PDB files alone with some external graphics viewer, that does not indicate whether or not the PSF has the S-S bond. If the HG1 atoms are still present, then it is NOT cross-linked.

Rotation alone does not undo a bond; a new specific patch would be needed to undo an S-S bond, and add back the H atoms. However, since the default CYS residue is not cross-linked, the only S-S bridges will be those explicitly created via the DISU patch. Suggested reading:

struct.doc
rtop.doc
top_all22_prot.inp

Read a textbook on simulations and the CHARMM papers.

If the two gamma sulfurs are close enough to form the S-S bond I assume that you would NOT want them to have the hydrogens still present.

You need to talk more to your supervisor - the forums are NOT a substitute for a supervisor! If your supervisor is clueless about this issue, well ....