Re: [AMBER] Fwd: Implicit solvent GB calculation

From: Carlos Simmerling <>
Date: Mon, 7 Mar 2011 10:10:41 -0500

you say you are using "Amber" but you don't say which program. For large
systems, pmemd is much more efficient than sander. Also keep in mind that
the tutorial options selected may not be appropriate for production systems.
You should be looking for peer-reviewed and successful applications on
similar problems and follow their methods. Tutorials are usually intended to
help you learn the program and execution, and not to suggest best practices
in terms of force field and model choices, and so on. GB models have well
documented weaknesses (and benefits), and you need to think carefully about
the potential impact on your project.

On Mon, Mar 7, 2011 at 10:04 AM, Rubben Torella <>wrote:

> Dear Amber community,
> I'm trying to perform MD simulation in implicit solvent using Generalized
> Born approach on a protein-DNA complex.
> I followed the tutorial described here (
> )<
> >in
> order to set-up the calculation.
> We are talking about 7637 atoms in the complex.
> I launched the calculation on 64 processors (10 ns), but the calculation
> seems to me very slow (1.3 ns/day)
> Is there something wrong in this approach or is it normal to have this
> calculation speed?
> I attached you the input of the MD calculation, in order to know if there
> are mistakes or there is some flags that slow down the calculation.
> Thank you for your help
> Rubben
> _______________________________________________
> AMBER mailing list
AMBER mailing list
Received on Mon Mar 07 2011 - 07:30:02 PST
Custom Search