[gmx-users] Error in generated Replica Exchange statistics in log file

Mark Abraham mark.j.abraham at gmail.com
Fri Feb 27 23:34:05 CET 2015


On Fri, Feb 27, 2015 at 12:07 PM, Abhi Acharya <abhi117acharya at gmail.com>
wrote:

> Hello GROMACS users,
>
> I have a problem with the Replica Exchange statistics printed in log file.
> I ran a 25 ns Replica exchange on a cluster through a queuing system.
> However, due to insufficient wall time the job stopped abruptly around 24
> ns and no stats were printed in the log file. I then extended the run by
> 10000 steps using -nsteps  and provided additional -cpi flag for
> continuation. The simulation ran without errors but the replica exchange
> stats that were printed in the log files of each of the replica is as
> follows:
>
> Replica exchange statistics
> Repl  10 attempts, 5 odd, 5 even
> Repl  average probabilities:
> Repl     0    1    2    3    4    5    6    7    8    9   10   11   12   13
>   14   15   16   17   18   19   20   21   22   23   24   25   26   27   28
>   29   30   31
> Repl      .23  .45  .57  .12  .28  .20  .82  .46  .44  .50  .07  .74  .53
>  .46  .44  .51  .22  .56  .24  .56  .23  .82  .30  .28  .59  .54  .34  .53
>  .75  .29  .62
> Repl  number of exchanges:
> Repl     0    1    2    3    4    5    6    7    8    9   10   11   12   13
>   14   15   16   17   18   19   20   21   22   23   24   25   26   27   28
>   29   30   31
> Repl        1    2    3    0    0    2    4    2    2    2    0    4    3
>  2    2    4    1    4    0    3    2    4    1    0    2    4    2    1
>  4    1    2
> Repl  average number of exchanges:
> Repl     0    1    2    3    4    5    6    7    8    9   10   11   12   13
>   14   15   16   17   18   19   20   21   22   23   24   25   26   27   28
>   29   30   31
> Repl      .20  .40  .60  .00  .00  .40  .80  .40  .40  .40  .00  .80  .60
>  .40  .40  .80  .20  .80  .00  .60  .40  .80  .20  .00  .40  .80  .40  .20
>  .80  .20  .40
>
>
>
> I rate of exchange attempt used was 1000, which means that the above stats
> somehow only take into account the last 10000 steps of the simulations. Is
>

This is normal. The subsequent mdrun will not read from the old .log file,
and the data is not in the checkpoint.


> this some bug or I did something wrong ? More importantly, is there a way
> to regenerate the replica exchange stats from the log file ?
>

Not without writing a custom script to parse the per-exchange information.

Mark

>
> Regards,
> Abhishek Acharya
> Shasara Research Foundation
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>


More information about the gromacs.org_gmx-users mailing list