Hi,
Thanks for the files. So at first I thought there was a bug in
cpptraj, but it turns out that even though the trajectory has the
'cell_lengths' etc variables, they are empty (all zeroes)! This is
part of the output from 'ncdump -v cell_lengths minimization.mdcrd':
```
data:
cell_lengths =
0, 0, 0,
0, 0, 0,
```
The angles are also zero. Cpptraj sees this and thinks because all the
lengths and angles are zero, there's no box (because effectively there
really isn't a box). I should probably have a warning printed there.
So it seems the problem is actually somewhere in sander. If you know
the dimensions of your box (I think I remember your simulation was
NVT), you can add the box information back using cpptraj with the
'box' command.
Hope this helps,
-Dan
On Fri, May 22, 2020 at 9:59 AM emanuele falbo <falbo.emanuele.gmail.com> wrote:
>
> Dear both,
>
> Yeah, that's right. This is a trajectory from the minimization.
> cell_lengths variable is there when running "$ ncdump -h trajectory.mdcrd":
>
> double cell_lengths(frame, cell_spatial) ;
> cell_lengths:units = "angstrom" ;
>
> So, there is actually something wrong in reading the trajectory in cpptraj.
>
> Is there any way I can add the box to the trajectory.mdcrd, and thus start
> the MD with such new box?
> Also, will it generate unphysical issues, like self-interactions, due to
> the different PBC?
>
> Cheers,
> Manuele
>
> Il giorno ven 22 mag 2020 alle ore 13:32 David A Case <
> david.case.rutgers.edu> ha scritto:
>
> > On Fri, May 22, 2020, Dan Roe wrote:
> > >
> > >Is trajectory.mdcrd the trajectory file from the QM/MM simulation? Double
> > >check that box info is there with ‘ncdump -h’ and look for the variable
> > >cell_lengths. If it isn’t there then for some reason box info wasn’t
> > >written to the trajectory file. If it is there, cpptraj is missing the
> > info
> > >somehow.
> >
> > Just a note: this is a "trajectory" file from a minimization run, not
> > from MD. It could be that there is something odd about that code path,
> > which is rarely used, and probably not tested.
> >
> > ....dac
> >
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
>
>
> --
>
> *Emanuele Falbo*
> PhD student
> Penfold group
> School of Natural and Environmental Sciences
> Bedson Building, Newcastle University
> Newcastle upon Tyne, NE1 7RU
> w: http://tompenfoldresearch.weebly.com/
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri May 22 2020 - 09:00:02 PDT
