[gmx-users] Re: Re: g_bundle problem

[Justin A. Lemkul]

Stefan Hoorman wrote:
> 
> 
>     Stefan Hoorman wrote:
>      >
>      >     Stefan Hoorman wrote:
>      >      > I have tried using g_bundle in order to analyse helix axes
>     in my
>      >      > transmembran helices. I created two groups in my ndx file that
>      >     included
>      >      > the alpha carbons of the first half of my helix and a second
>      >     group for
>      >      > the alpha carbons for the second half. When I try using
>     g_bundle the
>      >      > following error comes up:
>      >      > "Fatal error:
>      >      > The size of one of your index groups is not a multiple of n"
>      >      > What am I doing wrong?
>      >
>      >     Not a clue - What is your exact command line?  How many atoms
>     are in
>      >     each index
>      >     group?  The value of n is the number of parts into which the axes
>      >     are divided
>      >     and there are several criteria in the code that must be met, else
>      >     you get this
>      >     error.  So without this information, the best guess is: some
>      >     criterion is not
>      >     being met.
>      >
>      >     -Justin
>      >
>      >      > Thank you
>      >      >
>      >
>      >     --
>      >
>      >     g_bundle analyzes bundles of axes - and therefore number of atoms
>      >     each group
>      >     have to be divisible by number of axes (entered by -na). And I
>      >     suppose, you
>      >     are interested in analyzing only one axis, don't you?
>      >
>      >     Milan
>      >
>      >
>      > Well, the exact command line was : g_bundle -f md.xtc -s md.tpr -z -n
>      > bundle.ndx
>      > As for analyzing one axis, yes, that is true, I am only interested in
>      > analysing my helix principal axis.
>      > I have tried using different group sizes for the index group. I have
>      > tried using only carbon alpha for the first half and then for the
>     second
>      > half (each had 15 alpha carbons), then I tried with all sorts of
>      > different groups, first five residues and last five residues, first
>      > fifteen residues and so on.
>      > All analysis came up with the same error.
>      >
> 
>     You're not setting the -na flag, which by default is zero.  In the
>     code, if the
>     value from -na (int n) is <= 0, you get the fatal error.  Maybe the
>     error
>     message should be made more specific, so that users actually know
>     what n is.
>     You need to explicitly set a value for -na >= 1.
> 
>     -Justin
> 
> 
> Right, thank you Justin, things worked out nicely now. I was able to 
> calculatemy helix tilt with respect to the z axis. Would there be a way 
> to calculate the tilt between two helix axes? I cannot simply calculate 
> the tilt of each with respect to the z axis and subtract because the 
> interhelical tilt does not depend only on the z axis. Can g_bundle give 
> me that kind of information?
> 

It seems like it should:

http://lists.gromacs.org/pipermail/gmx-users/2007-October/030415.html

There are other programs that calculate angles, like g_sgangle and g_angle, both 
of which could potentially be used to calculate what you want.

-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

========================================