[gmx-users] check point file corrupted - umbrella sampling

[Poojari, Chetan]

Hello Chris,

I followed the steps as mentioned in the umbrella sampling tutorial in gromacs.

After NPT equilibration step, i did the umbrella simulation runs. Below are the commands which i used:


grompp -f md_umbrella.mdp -c npt0.gro -t npt0.cpt -p topol.top -n index.ndx -o umbrella0.tpr

mdrun -pf pullf-umbrella0.xvg -px pullx-umbrella0.xvg -s umbrella0.tpr -o 0.cpt -c 0.gro -x 0.xtc


So, I had 24 such commands in my job submission script (one script i had all the grompp command and in other script i had mdrun command) for 24 windows, which  gave me 24 cpt files.

After the  final 24th window run, along with the output names specified by me it also outputted default gromacs file names state_prev.cpt, state.cpt, md.log, ener.edr.

Please can I know if its necessary to output even log files for these 24 windows.

Size of the  state.cpt or the other cpt files which i got during npt equilibration are 1.2 MB (these cpt files are okie when i do gmxcheck, it does not give any error), where as of 24 different cpt files, each has file size of 715MB. I had run simulations for 10ns on each window.


Kind Regards,
chetan









________________________________________
From: gmx-users-bounces at gromacs.org [gmx-users-bounces at gromacs.org] On Behalf Of chris.neale at utoronto.ca [chris.neale at utoronto.ca]
Sent: 20 September 2011 13:37
To: gmx-users at gromacs.org
Subject: [gmx-users] check point file corrupted - umbrella sampling

I am somewhat confused. When I do umbrella sampling, I use a bash
script to set up a number of separate restrained simulations. Thus I
would get one .cpt file for each simulation. What are you doing that
in addition to these .cpt files you also get a state.cpt and
state_perv.cpt? Perhaps you had better start at the beginning and
explain what you did.

Nevertheless, if your .cpt file is indeed corrupt, then you can not
use it. You must extract a .gro from the final frame and use it (along
with your .trr and .edr if you like) to generate a new .tpr file that
will start a new run from where you left off.

Chris.

-- original message --

Hello Chris,


Thank you very much for your reply. I have done following test as you
suggested to check whether the checkpoint file are corrupted.


1) Yes, I have used  same version of gromacs for mdrun and to produce
.cpt file.


2)  when i do gmxcheck -f  0.cpt         # i get the same error.

Fatal error:
Start of file magic number mismatch, checkpoint file has 1993, should
be 171817
The checkpoint file is corrupted or not a checkpoint file


3)  When I do mdrun -s umbrella0.tpr       # it runs without any error.



4) My mdruns has not outputted anything like 0_prev.cpt, i just have 0.cpt.


I had run US for 24 windows, each for 10ns and produced 24 different
cpt files. Along with these 24 different .cpt files, in the end it
also produced state.cpt and state_prev.cpt.

When i did gmxcheck -f state.cpt    # i get

Checking file state.cpt

# Atoms  51895
Last frame         -1 time 10000.000


Item        #frames Timestep (ps)
Step             1
Time             1
Lambda           1
Coords           1
Velocities       1
Forces           0
Box              1


When i did gmxcheck -f state_prev.cpt   # i get

Checking file state_prev.cpt

# Atoms  51895
Last frame         -1 time 9730.730


Item        #frames Timestep (ps)
Step             1
Time             1
Lambda           1
Coords           1
Velocities       1
Forces           0
Box              1


I just have only one md.log file in the end, i haven't outputted
md.log file for each window.




Kind Regards,
Chetan.


________________________________________
From: gmx-users-bounces at gromacs.org [gmx-users-bounces at gromacs.org] On
Behalf Of Chris Neale [chris.neale at utoronto.ca]
Sent: 19 September 2011 17:59
To: gmx-users at gromacs.org
Subject: [gmx-users] check point file corrupted - umbrella sampling

I have used tpbconv -until to extend a US simulation without error,
and I assume that tpbconv -extend is also fine for US simulations.

1. Are you using the same version of gromacs mdrun as you used to
produce your .cpt file initially?

2. what happens when you run gmxcheck on the .cpt file?

3. what happens when you run mdrun with the umbrella0.tpr file?

4. what happens when you use 0_prev.cpt?

What I am getting at is that you have stated that the checkpoint file
is corrupted in your subject line, but you have not provided any
evidence of that. I do appreciate that you have pasted your exact
commands. The next step is to see if there is any way to test your
hypothesis that the checkpoint file is corrupted.

Chris.

-- original message --


Hi,

I am using Gromacs 4.5.3 and want to extend my umbrella sampling runs.

But i get an errror while running the "mdrun" step of extension
command, saying "Start of file magic number mismatch, checkpoint file
has 1993, should be 171817. The checkpoint file is corrupted or not a
checkpoint file"


Below are the commands I have used to run the umbrella sampling run


grompp   -f md_umbrella.mdp    -c npt0.gro    -t npt0.cpt    -p
topol.top    -n index.ndx    -o umbrella0.tpr


mdrun    -pf pullf-umbrella0.xvg    -px pullx-umbrella0.xvg    -s
umbrella0.tpr    -o 0.cpt    -c 0.gro    -x 0.xtc


I used the below commands for extension:


tpbconv  -s  umbrella0.tpr   -extend 20000   -o umbrella_new.tpr

mdrun   -s umbrella_new.tpr   -cpi 0.cpt   -pf pullf-umbrella_new.xvg
  -px pullx-umbrella_new.xvg   -o new.cpt   -c new.gro    -x new.xtc
(at this step i get the error)


I have not touched the cpt files.  Its the same as it was outputed.


Please can I know what might have gone wrong.



Kind Regards,
chetan



--
gmx-users mailing list    gmx-users at gromacs.org
http://lists.gromacs.org/mailman/listinfo/gmx-users
Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-users-request at gromacs.org.
Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------
Forschungszentrum Juelich GmbH
52425 Juelich
Sitz der Gesellschaft: Juelich
Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
Vorsitzender des Aufsichtsrats: MinDirig Dr. Karl Eugen Huthmacher
Geschaeftsfuehrung: Prof. Dr. Achim Bachem (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
Prof. Dr. Sebastian M. Schmidt
------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------