[gmx-users] How to calculate the spatial distribution function for the last 50ns of the trajectory?

[Christian Blau]

Hi Peiyin,


There is a related redmine entry that suggests this might be fixed in newer versions of GROMACS 
https://redmine.gromacs.org/issues/2189

(from GROMACS2018 on). If you are using a newer GROMACS version, to rule out that this is an issue related to .tng, can 
you try using the older .xtc format instead?


Best,
Chrstian

On 2019-12-05 03:13, Peiyin Lee wrote:
> Dear users,
>
>     I have a simulation that has a 400 ns trajectory and I would like to
> calculate the spatial distribution function of the last 50ns. I followed
> the instructions on gmx_spatial manual and everything was ok if I run the
> first 50ns of the trajectory instead of the last 50 ns. However, once I
> changed the command from gmx_mpi trjconv -s md.tpr -f md.xtc -b 0 -e 50000
> -o b.tng -boxcenter tric -ur compact -pbc none
> to gmx_mpi trjconv -s md.tpr -f md.xtc -b 350000 -e 400000 -o b.tng
> -boxcenter tric -ur compact -pbc none
> and then ran gmx_mpi trjconv -s md.tpr -f b.tng  -o c.tng -fit rot+trans
> I would get an error saying:
>
> WARNING: Incomplete frame: nr 0 time 0
>
>
> WARNING no output, last frame read at t=0
>
> I have checked my trajectory with gmx_check and the trajectory is fine. I
> wonder why those two commands only work with -b 0, but not -b anyothertime.
> Does anyone have an idea? Thank you in advance!
>
> Regards,
> Peiyin