Re: water box size

From: Roman Petrenko (
Date: Sun Nov 08 2009 - 10:52:03 CST

if you want to know _minimum_ water box size for a protein with
certain number of residues in _any_ (not physical) conformation, i'd
suggest to run simulations in vacuum. The protein will collapse into
some conformation to satisfy intra-protein bacbkone and sidechain
hydrogen bonds. Then, computing minimum, maximum coordinates (using
VMD "measure" command) will give you a _rough_ estimate of protein

Once again, the conformation from vacuum simulation has nothing to do
with "folded" (or native) structure of the protein. This approach is
_only_ to get insights into _approximate_ system size.

P.S. For the maximum box size estimation you can use a power law for
radius of gyration, Rg, of intrinsically disordered proteins that
simply depends on the number of residues, Rg[A] = 1.93 N^0.598 (Kohn,

On Sun, Nov 8, 2009 at 8:57 AM, Joshua Adelman <> wrote:
> The size of the water box should depend on a couple of factors, and it is
> probably better to think about the size of your protein in terms of its
> spatial extent rather than the number of atoms (Imagine for instance a long
> extended unfolded protein vs its compact folded form - same number of atoms,
> but with different requirements on box size to fully solvate them).
> Next you have to think about what is going on with the periodic boundary
> conditions  and the cutoffs you are using. You don't want your solute
> interacting with its periodic image.
> It's probably worth taking a look at some basic textbooks on molecular
> dynamics if you aren't familiar with these concepts.
> Best wishes,
> Josh
> On Nov 8, 2009, at 8:23 AM, gurunath katagi wrote:
> Hi everyone..
> Is there any criteria on as to what must be the size of the protein taking
> the no. of residues and atom count ..
> in the tutorial of ubiquitin, the atom count is around is 600 and have used
> only 5 A water box so as to suit the learning requirement for a beginner..
> But in practical, say if you have protein with some 600 residues with 9000
> atoms, what will be the size of the protein ...
> i.e is there any criteria in choosing the size of the water box for a
> particulate no. of atoms or is it just based on the criteria that water box
> must be suffficeintly large enough to
> immerse the protein throughout the simulation...
> thank you
> --
> Gurunath M Katagi
> Graduate Research Student
> Bio molecular computation Laboratory
> Supercomputer Education & Research Center
> Indian Institute Of Science, Bangalore- 560012
> Karnataka State, India
> Phone:080-22932737-319,080-22933061
> email:

Roman Petrenko
Physics Department
University of Cincinnati

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:53:28 CST