Re: Waterbox used in minimization?

From: Giacomo Fiorin (
Date: Thu May 13 2010 - 16:08:13 CDT

Hi Cem, if you're not going to do anyway some equilibration MD runs with an
explicit solvent (atomistic or CG), the explicit water box loses its

When you set up a system with explicit water, you should always restrain or
fix the protein, and do some MD run to give the possibility to the water
molecules to properly solvate the protein. For example, to fill those gaps
that couldn't be filled when you set up the system (since the most common
way to set up a protein/water box is to remove waters that are in a steric
clash with the protein, there will be gaps). Only after that, you may move
the protein, by minimization or MD.

If you don't want to do that because you need a very cheap and quick method
to minimize protein structures, you should probably avoid the explicit water
box altogether (having that will slow you down, indeed), and look into an
implicit water model. It cannot replace an explicit water box, but it's a
first approximation at least. Much better than minimizing a protein in
vacuo. One way to do that with NAMD is couple it with APBS, but I have no
direct experience with how well they can play together. Google NAMD and
APBS together for more info.


---- ----
 Dr. Giacomo Fiorin
 ICMS - Institute for Computational Molecular Science - Temple University
 1900 N 12 th Street, Philadelphia, PA 19122
---- ----

2010/5/13 Cem Meydan <>

> Dear NAMD users,
> I'd like to use NAMD to perform minimization on some modeled proteins.
> However I'm curious if the minimization procedure requires a water box,
> whether NAMD uses interactions with water during the energy calculation in
> the CG method. I'll not perform equilibration/MD runs afterwards, so it
> would be strictly for minimization. Should I omit the solvation/setting up
> PBC steps? Would it affect the quality or the speed of the runs in any way?
> Thanks.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:54:07 CST