On Wed, Aug 27, 2014, Nadia Li wrote:
>
> I am running amber on a protein-ligand complex. The ligand binding site is
> a rather flat, exposed surface, and thus in the production run, after about
> 10ns, the ligand moved away from the site. To avoid this, I am thinking to
> add a small harmonic restraint (like 1 kal/mol?) to the ligand in the
> production run.
If you restrain the ligand, then it indeed won't move much. But remember that
motion is relative, so the protein could just drift away from the (now
stationary) ligand, and you would end up with essentially the same behavior
you have now.
To artifically keep the ligand in the binding pocket, you would need to devise
some sort of internal restraints between the ligand and protein. (In Amber
jargon, these are called "NMR" restraints, but they need not have anything to
do with NMR.) But, as is always true, you need to think about what you want
to learn from the simulation. Adding artificial terms to force a certain
behavior can be useful in many cases, but you need to have a strategy for
getting out of the simulation more than you "put in" in terms of assumptions.
....good luck....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Aug 27 2014 - 19:30:03 PDT
