[gmx-users] g_mindist -or inconsistencies with atom-pairs

[Ran Friedman]

Dear Shay,

What do you get when -or is present? Do the atoms always belong to the
same residues?
I suspect that since the calculation of the minimal distance is made for
all residues, what you get at the end is the atoms at minimal distance
between the last two residues. This seems like a bug and I suggest you
submit a bugzilla.

As a side note I think that the output will be given only for the last
two groups in case there are more than two. You may want to check this
as well if you submit a bug report.

It should be too difficult to fix, but would involve changes in a few
places in the code.

Best regards,
Ran.

shayamra at post.tau.ac.il wrote:
> Dear Gromacs users,
>
> We used g_mindist analysis as follows:
> g_mindist -f *.xtc -s *.tpr -o atom-pairs.out -od mindist.xvg -or
> res_mindist.xvg
>
> and surprisingly in the atom-pairs.out only several atoms in one
> residue (from group 1) were at minimum distance from group 2
> throughout the entire simulation. Checking this output manually showed
> it to be incorrect.
> res_mindist.xvg seems to hold the correct data.
>
> When running the same analysis, only omitting the -or flag, we get a
> *correct* atom-pairs.out.
> g_mindist -f *.xtc -s *.tpr -o atom-pairs.out -od mindist.xvg
>
> Comparing mindist.xvg from both scenarios shows no difference (files
> are identical).
>
> Is the -or flag supposed to affect the other output files of the
> g_mindist function?
>
> Supplementary details:
> 1. The same thing happened when g_mindist -s was supplied with either
> .gro, .pdb or .tpr files.
> 2. Index-groups 1 and 2 consisted upon full molecules with no
> ommisions (full protein and not only C-alpha).
> 3. We ruled out PBC by using -nopi
> 4. This is common with the following Gromacs versions: 3.3.3., 4.0.5,
> 4.0.7.
>
> Regards,
> -Shay
>
> --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/search before
> posting!
> Please don't post (un)subscribe requests to the list. Use thewww
> interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>