> Is sander able to simulate covalently bonded periodic infinite crystals?
No. I can't explain exactly why though.
> (I'm under the impression that in Bongkeun's graphite example,
> "bondbydistance" was occasionally inserting bonds between atoms on
> opposite sides of the unit cell,
There is no cell until you solvate, and even if you drew the box
first with setBox, bondbydistance would not know it (speaking as
the author). It sounds like there is a bug in bondbydistance if it
made bonds longer than the cutoff of 1.5 you used.
> and that is is why the system
> appeared to be curving in on itself during equilibration. Stability
> appeared to be restored when these bonds were cut. However I imagine
> that we might want to include bonds for the boundary atoms. It's not
> hard to add these additional bonds manually using the "bond" command
> in leap, but I suspect that sander will misinterpret them.)
They won't be crossing the added water box.
> Similar issues also come up when simulating infinite helices or long
> cellulose chains.
Infinite yes, but I don't see it for merely long chains.
Bill
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Oct 21 2010 - 20:30:03 PDT
