2015-02-24 9:52 GMT-07:00 Vitalie Botan <[email protected]>:
> Dear Christoph,
>
> I've just realized that the newest development version of csg_map (version
> 1.3-dev hgid: 78c98aa45ba2) produces quite odd results when reading Lammps
> dump files:
>
> when I write the trajectory in wrapped coors, i.e.
>
> ITEM: TIMESTEP
> 50000
> ITEM: NUMBER OF ATOMS
> 576
> ITEM: BOX BOUNDS pp pp pp
> -16.6546 16.6546
> -16.6546 16.6546
> -16.6546 16.6546
> ITEM: ATOMS id mol type x y z
> 1 1 1 1.48239 -14.3129 4.66771
> 2 1 1 1.10847 -12.8235 4.62104
> 3 1 2 1.98531 -12.2345 4.39821
> 4 1 2 0.408144 -12.8184 3.77732
> 5 1 2 2.11519 -14.4292 5.57303
> 6 1 3 2.26613 -14.6418 3.36911
> 7 1 4 1.76513 -14.4093 2.27462
>
> the coordinates of the bead corresponding to these 7 atoms as written by
> csg_map are:
>
> ATOM      2   B1 UNL     1      16.348 -139.948  37.050
>
> As you see there is a factor of 10 at least in the COM coordinates.
>
> I've check it with scaled coordinates dump:
>
> ITEM: TIMESTEP
> 50000
> ITEM: NUMBER OF ATOMS
> 576
> ITEM: BOX BOUNDS pp pp pp
> -16.6546 16.6546
> -16.6546 16.6546
> -16.6546 16.6546
> ITEM: ATOMS id mol type xs ys zs
> 1 1 1 0.544504 0.0703018 0.640133
> 2 1 1 0.533278 0.115015 0.638732
> 3 1 2 0.559603 0.132698 0.632042
> 4 1 2 0.512253 0.115167 0.613402
> 5 1 2 0.563502 0.0668092 0.667313
> 6 1 3 0.568033 0.0604275 0.601147
> 7 1 4 0.552992 0.067407 0.568288
>
> and it looks even worse in this case:
> ATOM      2   B1 UNL     1     182.893  26.598 203.596
>
> The only meaningful output I found for unwrapped coordinates:
>
> ITEM: TIMESTEP
> 50000
> ITEM: NUMBER OF ATOMS
> 576
> ITEM: BOX BOUNDS pp pp pp
> -16.6546 16.6546
> -16.6546 16.6546
> -16.6546 16.6546
> ITEM: ATOMS id mol type xu yu zu
> 1 1 1 34.7915 18.9962 4.66771
> 2 1 1 34.4176 20.4856 4.62104
> 3 1 2 35.2944 21.0746 4.39821
> 4 1 2 33.7173 20.4907 3.77732
> 5 1 2 35.4243 18.8799 5.57303
> 6 1 3 35.5752 18.6673 3.36911
> 7 1 4 35.0742 18.8998 2.27462
>
> with csg_map output
> ATOM      2   B1 UNL     1      33.906  21.495   7.762
>
> I am quite sure that the version from Nov 2014 had no problems with Lammps
> dumps in wrapped coordinates (I have even visualized it VMD drawing
> atomistic and cg trajectories over each other), so it should be a recent
> change in the source code.
> Is it an angstrom/nm convention issue or something else?
Yes, VOTCA uses nm internally and only converts to angstrom in
necessary (e.g. pdb output).
Also VOTCA assumes that the coordinates in LAMMPS dump files are in
nm, this is mainly due to the fact that the dump files have no
atom_style block in them to determine the unit system. This might be
where your factor 10 comes from, you can check that by converting to a
gro file, which uses nm and is fairly human-readable, instead. I am
not sure where else there could be an issue as the code of the
pdbwriter/lammpsreader hasn't been touched for about a year.

Concerning the LAMMPS reader, there is not much difference between
reading different coordinate systems. Scaled coordinated are
multiplied with the box length when read to make them into normal
coordinates. VOTCA's mapper can deal with wrapped as well as unwrapped
coordinates, because mapping are evaluated with respect to the 1st
atom of the mapping. However, looking at the code in lammpsreader.cc
it seems unwrapped coordinates (xu,yu,zu) are not actually processed,
I fixed that in the code:
<https://code.google.com/p/votca/source/detail?r=abd839eebfdf1ab9c409d399039e68212a043a3a&repo=csg>
All three coordinate systems should give the same result for pbc.

Christoph


> If you need I can provide xml and pdb files for a single snapshot.
>
> Best,
>
> Vitalie
>
>
> --
> You received this message because you are subscribed to the Google Groups
> "votca" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To post to this group, send email to [email protected].
> Visit this group at http://groups.google.com/group/votca.
> For more options, visit https://groups.google.com/d/optout.



-- 
Christoph Junghans
Web: http://www.compphys.de

-- 
You received this message because you are subscribed to the Google Groups 
"votca" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/votca.
For more options, visit https://groups.google.com/d/optout.

Reply via email to