[gmx-users] Re:gmxdemo: peptide to real protein and boom

David Mathog mathog at caltech.edu
Wed Feb 15 19:46:41 CET 2006

So I changed the message to this form, which I thought
was a  bit clearer:

  Making "rest" group for XXX containing NNN elements

where XXX and NNN are variables.  It also was made to dump the
atoms in the group.  As an earlier poster had noted, the first time
this message was emitted it was triggered by a single CA.  However,
this message was also emitted for XXX = 

  Acceleration, Freeze, Energy Mon., VCM, User1,
  User2, XTC, Or. Res, and QMMM.  

and in each of those cases NNN was 10874, which seems to
correspond to all of the atoms.

I tried removing the CA from the input PDB file and that did
eliminate the first "rest" message but all the others remained.
It did this with either gromacs or Amber ff files.

The "demo" script was allowed to complete and it ran apparently
normally.  At least it didn't blow up anywhere and the trajectory
didn't do anything untoward, like exploding or unfolding the protein.

Presumably these "rest" groups appear because I didn't specify
members for these fields.  But no members for at least some fields
would seem to be the normal state of affairs.  The ref_t/tau_t
check only occurs for the first "rest" group though, so taking out
the one CA removes that message.

Which brings us back to the default state of affairs.  Most PDB files
have a few ions in them.  Apparently putting them into "rest" is
somehow a bad thing (not clear to my why though.) So what exactly
would be the problem in putting these (ions) into the "sol" group,
if that exists, or some other group if it doesn't?  With proper
warnings, of course.

Again, this is a bit of a user interface issue.  Pretty much all 
PDB files have some ions so the default behavior of these programs
should be to handle those atoms gracefully rather than requiring
a manual intervention following a cryptic error message.


David Mathog
mathog at caltech.edu
Manager, Sequence Analysis Facility, Biology Division, Caltech

More information about the gromacs.org_gmx-users mailing list