[gmx-users] Dihedral PCA documentation

Mark Abraham mark.j.abraham at gmail.com
Tue Feb 28 17:18:01 CET 2017 

Hi,

On Tue, Feb 28, 2017 at 5:51 AM Anna Vernon <lappala.anna at gmail.com> wrote:

> Thank you Justin,
>
> that makes sense, thank you for help. However, I am struggling with the
> rest, due to lacking documentation & examples. Like Alex said, this
> web-page http://www.gromacs.org/Documentation/How-tos/Dihedral_PCA is
> outdated, but there is no new version of it available...


Well, when you figure out some improvements, please suggest them.
Unfortunately it is often the case with such community projects that
someone writes some code, contributes it, and then disappears leaving
nobody quite sure what it is good for or how it works :-)


> Or at least I
> am not finding it (and I have noticed that people on the list have asked
> questions similar to mine, but there were no answers to any of my
> questions, sadly).
>
> mk_angndx -s h3_md.tpr -n h3_ANGLE.ndx -type dihedral
>
> is what I now have done, and my next steps should be the following, I
> guess,
>

Paraphrasing, the PCA howto at your link says "1. use gmx mk_angndx, 2. use
gmx angle -or dangle.trr, 3. make covar.ndx 4. gmx trjconv -f dangle.trr"

trjconv -s h3_md.tpr -f dangle.trr -o resized.gro -n covar.ndx -e 0
>
> but I do not have a trr file because mk_angndx does not generate it
> (unlike the old documentation command angle)


mk_angndx never did that. You seem to be skipping the gmx angle command
that makes that trr file.


> and again, my tpr file
> contains waters, not just the molecule of interest that I have in my
> .gro file. I also do not understand the difference between the
> dangle.ndx file and covar.ndx file... What is resized.gro?


It's a hack that copes with the way somebody thought it was a good idea to
abuse the trr format for storing non-trajectory data, to suit things like
covariance analyses. Steps 3 and 4 seem fairly easy to follow, while I
agree that the reason for doing them is mysterious. :-)


> It would be
> really helpful to write out those explanations in documentation and
> perhaps give a more realistic example because this really becomes
> frustrating.
>

I'm open to suggested improvements, but a "realistic" example with 243
dihedrals seems unlikely to make the instructions easier to follow...

Mark

Thanks for your time & help, it is very much appreciated.
> Anna
>
>
>
> On 27/02/17 19:58, Justin Lemkul wrote:
> >
> >
> > On 2/27/17 9:55 PM, Anna Vernon wrote:
> >> Thanks Justin,
> >>
> >> but I did use this command, which is supposed to be dihedral:
> >>
> >> g_angle -f h3_nowater.gro -n h3_ANGLE.ndx -or h3_ANGLE.trr -type
> >> dihedral
> >>
> >
> > Sure, but g_angle is only going to give you the options presented to
> > it in the index file.  If those groups were created incorrectly (note
> > my comments were about mk_angndx, not g_angle) then you have to create
> > them properly.  You can't get dihedrals from valence angles :)
> >
> > -Justin
> >
> >> thanks
> >>
> >> A
> >>
> >>
> >> On 27/02/17 19:35, Justin Lemkul wrote:
> >>>
> >>>
> >>> On 2/27/17 9:29 PM, Anna Vernon wrote:
> >>>> Thanks Alex.
> >>>>
> >>>>
> >>>> tried the new syntax, below the output of g_angle, which did not
> >>>> get any
> >>>> easier... Хрень осталась...)))
> >>>>
> >>>> Group     0 (UB_th=108.4_297.062f) has    36 elements
> >>>> Group     1 (UB_th=108.9_384.092f) has    30 elements
> >>>> Group     2 (UB_th=109.7_794.962f) has     6 elements
> >>>> Group     3 (UB_th=111.5_376.562f) has     6 elements
> >>>> Group     4 (UB_th=110.1_221.752f) has    36 elements
> >>>> Group     5 (UB_th=109.0_297.062f) has    33 elements
> >>>> Group     6 (UB_th=109.5_430.952f) has    36 elements
> >>>> Group     7 (UB_th=113.5_585.762f) has    12 elements
> >>>> Group     8 (UB_th=121.0_376.562f) has    12 elements
> >>>> Group     9 (UB_th=119.5_351.462f) has    24 elements
> >>>> Group    10 (UB_th=124.0_669.442f) has    12 elements
> >>>> Group    11 (UB_th=112.5_167.362f) has    12 elements
> >>>> Group    12 (UB_th=121.0_669.442f) has    15 elements
> >>>> Group    13 (UB_th=109.5_276.142f) has    30 elements
> >>>> Group    14 (UB_th=107.0_418.402f) has    12 elements
> >>>> Group    15 (UB_th=112.2_365.682f) has     6 elements
> >>>> Group    16 (UB_th=109.5_271.122f) has     6 elements
> >>>> Group    17 (UB_th=109.5_271.122f) has    12 elements
> >>>> Group    18 (UB_th=110.0_365.682f) has     3 elements
> >>>> Group    19 (UB_th=111.0_292.882f) has     6 elements
> >>>> Group    20 (UB_th=112.2_338.902f) has     3 elements
> >>>> Group    21 (UB_th=111.0_359.822f) has     6 elements
> >>>> Group    22 (UB_th=108.0_401.662f) has     6 elements
> >>>> Group    23 (UB_th=110.1_288.702f) has     6 elements
> >>>> Group    24 (UB_th=115.2_443.502f) has     6 elements
> >>>> Group    25 (UB_th=107.5_433.462f) has     6 elements
> >>>> Group    26 (UB_th=120.0_383.252f) has    12 elements
> >>>> Group    27 (UB_th=120.0_334.722f) has    36 elements
> >>>> Group    28 (UB_th=120.0_251.042f) has    60 elements
> >>>> Group    29 (UB_th=110.1_279.742f) has    18 elements
> >>>> Group    30 (UB_th=105.8_351.462f) has     3 elements
> >>>> Group    31 (UB_th=112.1_343.092f) has     6 elements
> >>>> Group    32 (UB_th=116.5_418.402f) has     3 elements
> >>>> Group    33 (UB_th=122.5_627.602f) has     3 elements
> >>>> Group    34 (UB_th=120.0_418.402f) has     6 elements
> >>>> Group    35 (UB_th=120.0_192.462f) has     3 elements
> >>>>
> >>>
> >>> These are Urey-Bradley angles with the given equilibrium angles and
> >>> force
> >>> constants.  If you're using mk_angndx to create the index group, you
> >>> want
> >>> "-type dihedral" instead of "-type angle" (which is the default).
> >>>
> >>> -Justin
> >>>
> >>>>
> >>>> On 27/02/17 16:04, Alex wrote:
> >>>>> I think you might be referring to the 2010 version of the online
> >>>>> manual (
> >>>>> http://www.gromacs.org/Documentation/How-tos/Dihedral_PCA), which
> >>>>> is pretty
> >>>>> outdated. A lot of syntax seems to be changing fairly rapidly, so
> >>>>> хрень
> >>>>> like this is to be expected. ;)
> >>>>> The current manual points to gmx angle, gmx covar and, gmx anaeig for
> >>>>> covariance analysis, and for those the updated help pages are
> >>>>> separate:
> >>>>> http://manual.gromacs.org/programs/gmx-angle.html
> >>>>> http://manual.gromacs.org/programs/gmx-covar.html
> >>>>> http://manual.gromacs.org/programs/gmx-anaeig.html
> >>>>>
> >>>>> Alex
> >>>>>
> >>>>> On Mon, Feb 27, 2017 at 3:44 PM, Anna Lappala
> >>>>> <lappala.anna at gmail.com>
> >>>>> wrote:
> >>>>>
> >>>>>> Dear Gromacs users,
> >>>>>>
> >>>>>> The documentation on Dihedral PCA analysis is confusing.
> >>>>>>
> >>>>>>   Firstly, the links are redirecting to the main gromacs page
> >>>>>> which does
> >>>>>> not help because one needs to search for the specific commands
> >>>>>> from there.
> >>>>>>
> >>>>>> Secondly, -s option does not work in step 2: invalid command line
> >>>>>> argument
> >>>>>> "-s".
> >>>>>>
> >>>>>> Thirdly, my tpr file contains water which I do not want to
> >>>>>> include in the
> >>>>>> analysis, how can I change that?
> >>>>>>
> >>>>>> .ndx file output produced by mk_angndx is a mess- first line, for
> >>>>>> example,
> >>>>>> is [UB_th=108.4_297.062f]
> >>>>>>
> >>>>>> I need a small number of dihedrals, so I could write the file
> >>>>>> myself but
> >>>>>> there again is no clear explanation how to do it: the [foo] 1 2 3
> >>>>>> 4 example
> >>>>>> could be expanded- do I use "dihedrals" instead of foo? Or
> >>>>>> filename? Or
> >>>>>> molecule name?
> >>>>>>
> >>>>>> Finally, I would really appreciate it if someone could give an
> >>>>>> explanation
> >>>>>> or a clear example of how this works, for a simple system or a
> >>>>>> protein,
> >>>>>> because documentation is unclear to me.
> >>>>>>
> >>>>>> Kind regards,
> >>>>>> Anna
> >>>>>> --
> >>>>>> Gromacs Users mailing list
> >>>>>>
> >>>>>> * Please search the archive at http://www.gromacs.org/
> >>>>>> Support/Mailing_Lists/GMX-Users_List before posting!
> >>>>>>
> >>>>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >>>>>>
> >>>>>> * For (un)subscribe requests visit
> >>>>>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users
> >>>>>> or
> >>>>>> send a mail to gmx-users-request at gromacs.org.
> >>>>>>
> >>>>
> >>>
> >>
> >
>
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.

More information about the gromacs.org_gmx-users mailing list