Re: [AMBER] A good way to use an Amber force field with NAMD?

From: Sasha Buzko <>
Date: Tue, 03 Mar 2009 16:26:44 -0800

to answer your question about the requirement: we are trying to get some
supercomputer time, and they only have NAMD installed. In any event,
it's not our choice.

With regard to the equilibration (comment by Gustavo), I suspect that
minimization/equilibration from NAMD would be equally valid. This has
more to do with the learning curve, since we need to get this working
correctly quickly. Unfortunately, NAMD doesn't seem to have as good a
set of tutorials and manuals as Amber does (thanks, Ross). They mention
the meaning of each keyword in ten different places, but never explain
the whole process or post a sample set of config files for the entire
simulation - minimization, equilibration, production run...

In other words, I'm worried that with this fragmented information about
NAMD setup, things can get messed up fairly easily. Hence, the interest
in using Amber to prepare the system and getting to NAMD just for the
production run...

BTW, if someone has a reference to a good description of the entire
simulation pipeline with NAMD, it would be very helpful (the official
NAMD manual and tutorials don't seem to have it)..



Adrian Roitberg wrote:
> Perhaps I can raise a different issue?
> I would like a clarification on your statement of: " We have a
> requirement to run some large simulations using NAMD". Is this because
> you have a contract with them or something ?
> Why not use amber (pmemd) directly ? I ask, because the usual comments
> about NAMD being much faster or scalable than PMEMD are usually not
> quite right.
> a.
> Gustavo Seabra wrote:
>> Assuming the NAMD implementation of the Amber force field is correct,
>> I don't see a reason for the two approaches to yield different
>> results. In fact, if you are going to (have to) trust the NAMD results
>> for the production run, why not trust them for the equilibration run
>> as well?
>> Gustavo.
>> On Tue, Mar 3, 2009 at 4:37 PM, Sasha Buzko <> wrote:
>>> Hi all,
>>> We have a requirement to run some large simulations using NAMD, and the
>>> system includes a bunch of nonstandard items that are not
>>> parameterized in
>>> the force field that comes with NAMD. As a solution, we are considering
>>> using Amber's prmtop and coordinate files.
>>> Now, we see two possible ways to do this: First: minimize and
>>> equilibrate
>>> using Amber, and then use the restart file of the equilibrated
>>> system for
>>> the production run with NAMD. Second: do the whole thing in NAMD. We
>>> are a
>>> bit more comfortable with using Amber part of the way, mostly due to
>>> more
>>> experience with it..
>>> The question is: are there any substantial differences between these
>>> two
>>> approaches? If the system is minimized and equilibrated in Amber,
>>> and then
>>> simulated with NAMD, will it affect the outcome? Can anyone comment on
>>> possible issues?
>>> Thanks in advance for any comments
>>> Sasha
>>> _______________________________________________
>>> AMBER mailing list
>> _______________________________________________
>> AMBER mailing list
AMBER mailing list
Received on Wed Mar 04 2009 - 01:27:07 PST
Custom Search