[gmx-users] REMD protein-ligand separation

[Mark Abraham]

David Osguthorpe wrote:
> 
> Hi,
> 
> I have performed a REMD simulation of a protein and ligand in water
> 
> as usual the protein and ligand get separated by symmetry operations
> 
> however attempts to fix this using trjconv and the standard recipes
> (eg. -pbc whole, -pbc nojump) do not appear to be working
> 
> in particular after using -pbc nojump  and attempting RMS -fit rot+trans
> I get warning messages
> 
> ....
> There were 26 inconsistent shifts. Check your topology
> Will stop reporting inconsistent shifts
> 
> (no such errors appear in the REMD logs)
> 
> is there any way to get the protein and ligand to be together?

There is some experience under 3.3.x that multi-stage trjconv is more 
effective than one-stage trjconv. I don't recall any details, but you 
may find some if you search the mailing list.

Mark