[gmx-users] Implicit solvent MD is not fast and not accurate.

[Justin Lemkul]

On Wed, Mar 27, 2013 at 10:11 AM, xiao <helitrope at 126.com> wrote:

> Dear Justin,
> Thank you very much for your reply.
> I found that the speed of implict MD is slower that explict MD. For
> examplex, the speed of an explict MD for a protein of 300 amino acids is
> about 3ns per day, however, the implicit solvent is about 1.5ns per day.
>

I know. This is what I mean about the need for improving the capabilities
of the code. The main point of the implicit solvent approach is that it
should be faster.


> With respect to the accuracy of implicit solvent, the result shows bad
> result. There are two carbon atom types which are not in the gbsa.itp, and
> i just copied some carbon atom type in gbsa.itp because i find that there
> is no big difference between the carbon parameters. I do not know whether
> this is the

reason.
>

You need to establish whether what you did was appropriate or not before
you begin accusing Gromacs of being inaccurate.  If you observe
inconsistencies, errors, poor energy conservation, etc. in the context of
some known system, then that might be worth investigating.  Problems with a
custom topology are more likely to be due to the topology than to the
software using it.  GB atom types are one possible source of error, but in
parametrization, there are plenty of things that can go wrong before you
even get to that point.

-Justin

-- 

========================================

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540)
231-9080http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================