AMBER Archive (2006)

Subject: RE: AMBER: how to combine NEB method with QM/MM approach

From: Ross Walker (ross_at_rosswalker.co.uk)
Date: Mon Oct 09 2006 - 17:33:00 CDT

Dear Caoch

> i want to study a proton transfer recation pathway using QM/MM-MD
> approach with NEB method.
>
> My questions are
> I) Can such calculation be done only by PIMD?
> if it is not, should such calculation be divided into two
> steps? (i.e.
> extract structures by NEB method and compute energy by QM/MM method)

Yes the calculation can be done in Amber 9. The executable you need to use
is sander.PIMD.MPI. This is really just due to coding reasons than to any
specific method being used. Thus while the exeutable is called PIMD the NEB
calculation is not related to PIMD and should not be confused.

I suggest that you take a look at the following two tutorials:

http://www.rosswalker.co.uk/tutorials/amber_workshop/Tutorial_seven/

and

http://www.rosswalker.co.uk/tutorials/amber_workshop/Tutorial_Six_amber9/

The first will explain how to run classical NEB simulations and the second
will show you how to run QM/MM simulations.

I suggest you start by making sure that you can create prmtop and inpcrd
files for your two end points and run stable QM/MM MD simulations on them.
You can then create an NEB simulation out of these.
 
> II)it seems NEB is used for studying conformation changes in
> AMBER. in
> the case of studing recation, how can I get good structures
> along recation
> pathway by NEB methods?Is it necessary to give enough initial
> structures
> before starting NEB calculation?

This is a tough one to answer and really the subject of research.
Technically you do not need to provide a pathway. You just set say 32 images
to be one configuration and the other 32 to be the other configuration.
Sander will handle this situation in the code and you will end up with
images being pulled along a coordinate between the two end points. Hopefully
if you carry out sufficient simulated annealing you will find a suitable low
energy reaction coordinate. However, you can reduce the amount of sampling
potentially required by creating input files that contain extrapolations
along what you think the reaction path might be. This is very tough to setup
by hand though so you might want to try the first approach first.
  
> III) Are there some references for such study done by amber9?

There is a conformational change published which is:

Mathews, D.H., Case, D.A., JMB 2006, 357, 1683-1693

Although this only used classical NEB to investigate a conformational
change. I have tested the QM/MM implementation with small systems to verify
that the code works and this was published as a paper at the Fall 2005 ACS
meeting in San Diego. However, I have not tested this method on 'real'
reaction pathways yet so I will be interested to see how it does.

Try out the two tutorials. These should give you enough pointers to get
going. If you find any strange behaviour please contact me.

All the best
Ross

/\
\/
|\oss Walker

| HPC Consultant and Staff Scientist |
| San Diego Supercomputer Center |
| Tel: +1 858 822 0854 | EMail:- ross_at_rosswalker.co.uk |
| http://www.rosswalker.co.uk | PGP Key available on request |

Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu