Topic Options
#11612 - 08/24/06 04:19 AM What is the difference between PBC and CPT
accelrys Offline
Forum Member

Registered: 06/21/06
Posts: 21
I just want to optimize a protein.I have no idea how to choose the method,PBC(Periodic Boundary Condition)MD simulation or CPT(Constant Pressure&Temperature)MD simulation. What is the difference between them? Thanks!

#11613 - 08/24/06 07:55 AM Re: What is the difference between PBC and CPT [Re: accelrys]
lennart Offline

Forum Member

Registered: 09/25/03
Posts: 4742
Loc: ~ 59N, 15E
You should try to get hold of an introductory book on simulations. There are also references in pressure.doc to the original literature.

PBC is a way of treating the boundary between your finite system.
CPT refers to the ensemble in which the simulation is performed; in this case it would be constant pressure and temperature (and number of particles). If you propagate your system with PBC using just Newton's equations of motion you get the microcanonical ensemble (constant volume and energy). By applying special algorithms it is possible to keep certain variables (eg temperature, pressure, surface tension) constant. Note that not all such methods correspond to a proper ensemble.
Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden

#11614 - 08/24/06 10:58 AM Re: What is the difference between PBC and CPT [Re: accelrys]
rmv Offline

Forum Member

Registered: 09/17/03
Posts: 8379
Loc: 39 03 48 N, 77 06 54 W
A couple additional notes--

* In general, constant P methods often require the use of PBC to impose a volume that can change in size in response to the pressure.

* CPT is also a CHARMM keyword for a particular algorithmic approach (extended system) that can be used to perform simulations in a variety of statistical mechanics ensembles, including NPT, NVT, NPH, and specialized anisotropic ensembles such as NPAT or NPgT for lipid bilayers and other systems with a planar interface.

* The CPT keyword requires the use of CRYSTAL for PBC

* I suspect no one knows exactly what you mean by "optimize a protein"; I know I don't.
Rick Venable
computational chemist


Moderator:  John Legato, lennart