CHARMM Development Project
Posted By: firer constraint for the center of mass of protein - 12/05/03 10:38 PM
Hi, CHARMMers,
I want to perform a MD simulation to investigate protein interaction with surfaces. The question I have is that If wanna to keep the distance between surface and protein as constant, I have to fix the center of mass of protein.
Can CHARMM do that? which command I should use?
or, to be simple, I can initially fix the atom which is very close to the center of mass of protein by using "CONS FIX " command during the simulation. I knew CHARMM can do it using the latter method. But, I don't know whether this method is reasonable ?

Thanks,
mmfp.doc contains a variety of examples of how to constrain distances with respect to centers of masses (you'll have to invoke the MMFP command prior to setting up the constraining/restraining potential). In addition, the cons.doc describes at least two different approaches (e.g. cons hcmc or dmco) for these types of operations.
Another solution is to look at cons.doc which describes several commands that can do this (in one way or another);

CONS HMCM
CONS DROP
NOE
RESDist
CONS FIX
CONS HARM

From the initial question, is seems that CONS HMCM might be the most logical. NOE is good for restraining a distance between two large selections. Finally, the lonepair facility can be used to create additional centers which can be used in a creative manner...
Actually I've been consuming a substantial amount of time in testing different approaches for rcm constraints including methods described in MMFP.doc or in CONS.doc. Some of the solutions presented in the latter case are rather new features- right?

The creation of a lone-pair - or ghost atom - at a point relative the molecule and then restrain the interatomic distances is a nice idea..

What is the problem with using MMFP?
I didn't say that anything was wrong with the mmfp code. My only real objection is that the documentation is rather obtuse.

I recommend that all of these be used with a bit of testing. A few "TEST FIRSt" commands on a subset of atoms is a good way to ensure that the code is functioning properly.
Ok,
Thanks, I'll keep that in mind since one field of interest involve interactions between multicomplex macromolecular assemblies and how they dissociate/associate. Another key issue of interest is to remodel proteins accordingly to their overall shape with a simpler description for the individual aa's and then use a simplified center-center potential to describe the assembly between a set of macromolecules - in a similar way to what Dr.Huan-Xiang Zhou does (Biophys. J (1997) ,73, 2441) - but with a another description of the potential.

As long things don't get too complicated I believe that we can do fairly accurate calculations, and besides that, it's fun too.
© CHARMM forums