[gmx-users] File editing - only one layer of water around a molecule

[Mark Abraham]

On 15/04/2012 1:15 AM, Lara Bunte wrote:
> Hi Justin
>
> The difference in my
>
> g_select -s molecule_in_water.pdb -select '"Close to ISO" resname SOL 
> and within 0.5 of group "ISO"'
>
> and your
>
> g_select -s molecule_in_water.pdb -select '"Close to ISO" resname SOL 
> and within 0.5 of resname ISO'
>
> is, that the last ISO is not in quotation marks. Could you please 
> explain?

The significant difference is that you refer to "group "ISO"" without 
defining that group (since it is not defined by default, unlike the 
example group named Protein), and Justin refers to the RESidue NAMEd 
"ISO" with "resname ISO".

As you would know if you had more experience with the Unix command line, 
quotation marks are sometimes necessary to help delineate what groups of 
words make sense where (e.g. in English
Mark said "Justin said do this"
differs from
"Mark said," Justin said, "do this"). Sometimes you can get away with no 
quotation marks around something if the purpose of the quotation marks 
is merely grouping and the thing is only one word long. So the selection 
string needs quotation marks around "Close to ISO" to group those words 
to create a new index group with that name, the command line needs 
quotation marks around the selection string to group all those words, 
but Justin's resname might/can survive without. Unlike a reader of the 
written word, the UNIX shell parser is not endowed with higher 
intellect, so one needs to use different quotation marks to handle 
nesting, or explicitly delineate the nesting... but that is a topic you 
can Google about at your leisure.

>
> With your command I got this time no error but I got other stuff that 
> I don't understand:
>
> 1.)
> I got this warnings:
>
> WARNING: masses and atomic (Van der Waals) radii will be determined
>          based on residue and atom names. These numbers can deviate
>          from the correct mass and radius of the atom type.

Unlike programs that just guess stuff and don't ever tell you about it 
(which explains 99% of the posts on this list about the behaviour of 
VMD), GROMACS warns you that it's going to have to make guesses about 
your input since it comes in .pdb form. Input in .tpr form would allow 
the tool to make definitive assignments. If you had an atom named "HG" 
that was intended to be mercury or a hydrogen on a gamma carbon, you 
should go check the results made sense. So you should judge for yourself 
how weird your atom names are with respect to standard PDB naming, and 
whether you even use the masses or VDW radii in your calculation.

>
>
> WARNING: if there are broken molecules in the trajectory file,
>          they can not be made whole without a run input file
>
> Can I ignore this or is this serious?

I've now improved this error message in the code. Molecules can be 
"broken" across periodic boundaries under some circumstances. Since 
you're using .pdb input, and GROMACS tools make no attempt to use the 
PDB format for atomic connectivity even if it is present, GROMACS is 
warning you that you need to consider this aspect yourself. Inspecting 
the relevant coordinates in the file or a visualization program and 
comparing them with the location of the periodic box might be necessary, 
depending how your input file was created. You will normally know things 
about your files that GROMACS cannot know.

>
> 2.)
> My output is one file called size.xvg. It contains
>
> # This file was created Sat Apr 14 16:55:56 2012
> # by the following command:
> # g_select -s molecule_in_water.pdb -select "Close to ISO" resname SOL 
> and within 0.5 of resname ISO
> #
> # g_select is part of G R O M A C S:
> #
> # Green Red Orange Magenta Azure Cyan Skyblue
> #
> @    title "Selection size"
> @    xaxis  label "Time (ps)"
> @    yaxis  label "Number"
> @TYPE xy
> # Selections:
> #   "Close to ISO" resname SOL and within 0.5 of resname ISO
> #
> @ view 0.15, 0.15, 0.75, 0.85
> @ legend on
> @ legend box on
> @ legend loctype view
> @ legend 0.78, 0.8
> @ legend length 2
> @ s0 legend "Close to ISO"
>       0.000  848.000
>
> What is the interpretation of this? What I need is a new pdb. file 
> that has to contain only the molecule ISO and the water layer SOL 
> around it. Is this possible with g_select?

No. g_select has defaulted silently to the behaviour for g_select -os, 
which you can read about in g_select -h. That's no good to you, of 
course. You need an index group that you can use with another tool (e.g. 
trjconv) to actually make the selection. (Future GROMACS versions will 
be able to do this in one step, but that nirvana yet awaits.) For the 
exercise, please look up the correct output option for this from 
g_select -h.

Mark

>
> Thanks for helping me.
> Greetings
> Lara
>
>
> p.s.
> I use yahoo. I tried to find an "answer to all" option or something 
> like that. Again after pressing answer the mail should go only to 
> Justin. I put the mailing list in cc. Is here someone using yahoo also 
> and know how to fix this?
>
>
>
>
> ------------------------------------------------------------------------
> *Von:* Justin A. Lemkul <jalemkul at vt.edu>
> *An:* Lara Bunte <lara.bunte at yahoo.de>; Discussion list for GROMACS 
> users <gmx-users at gromacs.org>
> *Gesendet:* 16:56 Samstag, 14.April 2012
> *Betreff:* Re: [gmx-users] File editing - only one layer of water 
> around a molecule
>
>
>
> Lara Bunte wrote:
> > I still got the problem. What is wrong in this command:
> >
> > g_select -s molecule_in_water.pdb -select '"Close to ISO" resname 
> SOL and within 0.5 of group "ISO"'
> >
> > In the pdb. file ISO is for the molecule and SOL for the water.
> >
> > Please help
> >
>
> The above command assumes "ISO" is a default group, like "Protein" or 
> something else.  You can make selections based on any arbitrary 
> residue name with something like the following:
>
> g_select -s molecule_in_water.pdb -select '"Close to ISO" resname SOL 
> and within 0.5 of resname ISO'
>
> Does that work?
>
> -Justin
>
> -- ========================================
>
> Justin A. Lemkul
> Ph.D. Candidate
> ICTAS Doctoral Scholar
> MILES-IGERT Trainee
> Department of Biochemistry
> Virginia Tech
> Blacksburg, VA
> jalemkul[at]vt.edu | (540) 231-9080
> http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
>
> ========================================
>
>
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20120415/e3c4addb/attachment.html>