Hi all,
I am trying to visualize the simulation of a multimeric protein in a octahedron box and I used several combinations of -pbc (whole, mol, no-jump) and -center but the waters are still not wrapped correctly. Is there any specific protocol to deal with the pbc wrapping of such systems?
Any help would be appreciated!
Hi! What exactly do you mean by “not wrapped correctly”? With an octahedral box, the gmx trjconv -ur compact option might be what you’re looking for.
Hi, when I use -pbc whole or -pbc mol (with centering around a residue in the center of the multimeric system), some parts of the protein go out of the box.
I also tried using -pbc nojump before and after the centering but in both cases it completely messes up the waters and its all over the place.
Ah okay, I’m not expert here but I would probably go with -pbc mol -ur compact -center then, using the entire protein for centering. That should prevent any weird jumps and make sure everything is wrapped correctly. You can center the view around a single residue later in the visualization software of your choice.
On top of what @lmullender suggested, then do another round with -pbc cluster to try to group the proteins back if you have PBCs issues. Run this after cleaning up with -pbc mol -ur compact -center, that is, on the output trajectory. Try also to use a tpr file as reference that is whole and close to a structure that you are trying to reconstruct (e.g. the one of energy minimization).
Hello, Thanks I used -pbc whole first, then -pbc cluster around the protein chains only (which I defined using a custom ndx) followed by -pbc mol and centering around the protein chains and it worked for 195/200 of the frames. Only in a few the chains are still separated in space. I found out -pbc nojump is the step to avoid, especially if the chain indices are not continuous. Thanks for the suggestions!