AMBER Archive (2007)

Subject: AMBER: RE: Amber query !!

From: Ross Walker (ross_at_rosswalker.co.uk)
Date: Mon Oct 08 2007 - 12:02:36 CDT

Hi Sampath,

I am copying this reply to the amber mailing list which is the best location
for these types of questions since answers end up in the mailing list
archive http://amber.ch.ic.ac.uk/archive/ - see http://amber.scripps.edu for
details on how to sign up.

> I don't understand the energy table:
>
> For eg:
>
> NSTEP ENERGY RMS GMAX
> NAME NUMBER
> 3000 -4.8998E+04 2.4877E+00 8.4881E+01
> HD2 2174
>
> BOND = 7438.7176 ANGLE = 437.6583
> DIHED = 1450.0382
> VDWAALS = 15576.8022 EEL = -79048.5290
> HBOND = 0.0000
> 1-4 VDW = 551.4244 1-4 EEL = 4596.1063
> RESTRAINT = 0.0000
>
> In this table:
> is energy: -4.8998E+04 the total energy of the system

Yes this is a minimization and -4.8998E+04 is the total energy of the system
in KCal/Mol. It is the sum of all the terms below such as bonds, angles etc.

> ?
> What is Gmax (maximum gradient ?)

Yes this is the maximum force on all the atoms. So this tells you which atom
has the highest force, in Kcal/Mol/A?. It is useful in minimization because
if you have a bad structure, say two atoms overlapping it will tell you that
one of the atoms has a huge force (e.g. might be E+08 or so)

> What is HD2 2174 ? (is it the Atom that has the
> maximum gradient ?)

Yes this is the atom to which GMAX corresponds. So if GMAX is very large
then this would give you an idea of where to look to find problems with your
initial structure.

> I also don't understand the time table:
>
> | Force time 1521.24 (100.0% of
> Runmd)
> | Runmd Time 1521.24 (98.38% of Total)
> | Other 25.03 ( 1.62% of Total)
> | Total time 1546.36 (100.0% of ALL )
>
> Which time should I consider for benchmarking purposes
> ?

For benchmarking you should look at the Runmd time. This is the total time
take for the run, in seconds, exluding initial startup time. The key point
here is that the initial startup time should remain largely constant
regardless of the length of the simulation. So Runmd gives you a better idea
of the time it will take to do nstlim's worth of md steps.

Good luck.
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