AMBER Archive (2009)

Subject: RE: [AMBER] the output files

From: Ross Walker (ross_at_rosswalker.co.uk)
Date: Tue Feb 10 2009 - 11:45:37 CST


Hi Oguz

to follow up on Carlos' comments.

> I examined the Amber file formats as you said and I still have some
> questions that I want to learn in details. Firstly, according to the file
> formats for 'restrt' version created as a result of energy minimization or
> molecular dynamics includes six colums. Could you explain the types of
> information ( coordinates, velocity and periodic box information ) of
> these
> colums ?

Exactly what part of the description do you not follow?

http://ambermd.org/formats.html

Coordinate/restart file specification
...
FORMAT(6F12.7) (X(i), Y(i), Z(i), i = 1,NATOM)
  X,Y,Z : coordinates
...

This is a simple Fortran format description. Go pick up a book on Fortran
and take a look at the section entitled Formatted I/O. So this says there
are 6 floating point numbers per line of 12 spaces by 7 decimal points and
they are written for X, Y and Z for 1 to natom. So this gives 6 entries per
line x1,y1,z1,x2,y2,z2 etc.

It then goes on to clearly describe the velocities and box information etc.
 
> Secondly for mdcrd, mdvel version created by molecular dynamics
> includes ten colums. I think Amber file formats have an information
> writing
> order. So, generally What are the types of information ( coordinates,
> velocity and periodic box information ) of these ten colums ?

If you plan to analyze / read the output files in your own code (which I
assume is why you need to know the formats etc) then I suggest you look
carefully at the AMBER source code (since it is kindly provided to you) this
will explain exactly what is in all the files since you will find the write
statements that create them.

E.g.

dynlib.f (subroutine corpac)
!+ Emit coordinates or velocities, r(istart:n), to unit nf.

This writes the mdcrd and mdvel files.

or

mdwrit.f (subroutine mdwrit)

This writes the restart files.

Also you should probably think back to the days when you learnt how to do
experiments in the lab. There is NO difference with computational chemistry
and I think it is a real shame that people seem to forget this. Why don't
you try some experimenting? You can go into the code and add print
statements. You can fire it up in a debugger and watch it execute line by
line - remember the word 'computational' in computational chemistry so if
you don't know how to do this sort of thing you should probably learn it
before trying to go any further. Also why not do some simple tests. Build
yourself a simple 4 or 5 atom structure and build the prmtop / inpcrd file
in gas phase. Then take a look at the inpcrd file and it should be obvious
to you that it runs x1, y1, z1, x2, y2, z2 etc. Then try running a single
step of MD. Then look at the restart and mdcrd files and since you know what
the coordinates should be you should be able to recognize everything.

> And I also
> asked a question to mail archive about analysis but I didn't get an answer
> to this question. Can I calcuate Mean square displacements ( MSD ) and
> autocorrelation functions of the stress tensors using ptraj or other
> analysis tools in Amber 10 ? I'm eagerly waiting for your suggestions.

If nobody replies this probably means you can't do it. Is it mentioned in
the manual? Have you tried searching for it and see if anybody mentions it.
Have people published Mean Square displacements and autocorrelations of
stress tensors in the literature? If so what programs do they say they used
to do it? Given the equations for these functions it should be fairly easy
to write something that does it but it helps to do an exhaustive search
first.

Remember that VERY BUSY people are donating their time to help here so you
should be prepared to put work in yourself before asking people to help you.

Good luck,
Ross

/\
\/
|\oss Walker

| Assistant Research Professor |
| 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.

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber