AMBER Archive (2006)

Subject: RE: AMBER: some puzzles in equilibration MD

From: Ross Walker (
Date: Mon Oct 09 2006 - 17:41:54 CDT

Dear Zhihong,

> In fact, I've minimized the system on step1 with 500 force
> constant to fix the solute and relax water only, then when I
> went on minimizing the whole system without restraint, the
> procedure couldn't complete because of very very high "1-4
> EEL". Therefore I checked the out file and found the strained
> atoms, it was a hydrogen atom of a hydroxy in a cofactor (a
> small organic molecule, ThDP), this cofactor has a net charge
> of +1, I generate its force field parameters with
> antechamber, gaussion 03 opt and RESP fit. Although I found
> this highly strained atom, what can I do then ?

There is not an easy way of yet to really fix such a clash. There are
several 'hacks' you can use. The first is to work out what atom number it
is. Then open the inpcrd file in a text editor and find the entry
corresponding to this atom. You can verify the inpcrd coordinates to those
printed in the info in a visualization program such as VMD to ensure you
have the correct one. Then try tweaking these numbers a bit to see if you
can improve the situation. This may get you to a structure that while still
strained can be improved to something reasonable with steepest descent

Alternatively you can write a pdb file from the prmtop and inpcrd files and
then load this into a program that allows you to manually move atoms by
hand. Then you can write this back as a pdb file, feed it into leap and
create a new prmtop and inpcrd file. Assuming of course that the
manipulation program you choose doesn't mangle the pdb format. Unfortunately
I don't know of any free programs that do this off hand. Perhaps someone
else on the list knows of some useful ones.

> Also, most of system are stable, but a tail of the ThDP with
> above highly strained hydrogen atom vibrated disorderedly

Yes this is the because that section has incredibly high forces due to the
clash. You could also try adding restraints to everything except the 4 atoms
or so involved in the 1-4 that gives you a high energy. Then run
minimization for 100 steps or so of steepest descent and then say 100 steps
of MD at very low temperature (say 10K) and a 0.2 fs time step (dt = 0.0002)
and no shake. This may allow that small part of the system to relax to a
more favourable configuration in controlled fashion. Make sure you set
ntwx=1 so that you can visualize every step and make sure the system soesn't
just blow up.

All the best

|\oss Walker

| HPC Consultant and Staff Scientist |
| San Diego Supercomputer Center |
| Tel: +1 858 822 0854 | EMail:- |
| | 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
To unsubscribe, send "unsubscribe amber" to