[gmx-developers] g_hbond development

[Erik Marklund]

Before I went on vacation I managed to make the necessary changes to
gmx_hbond.c. However, I forgot to tell you all about it (dunno if you're
interested at all, but here goes). I decided to go for a different
strategy than originally planned (see below). Here is what I wrote to
the log:

revision 1.16.2.16
date: 2006/07/03 14:25:08;  author: erikm;  state: Exp;  lines: +132
-110
Added option -r2 for providing a second cut-off (see my post at gmx-
developers about a weak ago)

I chose a different strategy than I originally planned (once again, see
gmx-developers). Instead, when the program runs in contact-mode,
contacts are stored in the h-bond array h[] and contacts within r2 are
stored in the contact array g[]. This saves a lot of memory and most
parts of the code was insensitive, or could be 'desensitized', to
whether h-bonds or contacts was to be analyzed. Note however, that
is_hbond(...) now returns hbHB for contacts when operating in contact
mode.

I haven't had time to bug test the new code much, but when I return from
my summer vacation the 1:st of August I'll do it more thoroughly. I
guess there may be
some confusing output saying 'contacts' instead of 'h-bonds' and the
other way around

Erik Marklund, PhD student
Department of cellular biology, Molecular biophysics group
Uppsala University
erikm at xray.bmc.uu.se

-----

There you go.
I still haven't had time to test it properly, but it won't be long now.

/Erik


On Mon, 2006-06-26 at 19:22 +0200, Erik Marklund wrote:
> Having recently made a bunch of changes in g_hbond (and comitted it to
> the cvs repository), I have discovered that the -ac option still isn't
> working properly in conjunction with the -contact option. The problem
> is not just a programming error, but actually insuficient
> theory/phonomenology. When analyzing h-bonds everything is fine.
> Forward and back rate constants are calculated for the reaction A <=>
> B, where A = 'h-bonded' and B = 'within cut-off but not h-bonded'. For
> contacts, however, A is 'within cut-off range R', but how should B be
> defined? One possibility is to use B = !A, which doesn't follow the
> theory being described in Luzar 2000 and that is used in the present
> code for h-bonds. Another possibility is to define a second cut-off
> range R2 and say that B = 'within R2, but not within R'.
>  
> I have posted a bugzilla (#88) and hope to resolve this issue all by
> myself this week. However, any suggestions or wise comments are more
> than welcome. This is the stratregy I have decided to go for:
>  
> - introduce option -R2, being the second cut-off (see above)
> - add an additional array to the t_hbond structure containing contacts
> within R2, similar to g[] (see code).
> - if -R2 is positive and -contact and -ac flags are set, calculate the
> autocorrelation stuff using the second definition of B above.
> - if R2 is zero, B is defined as !A and the autocorrelation stuff is
> calculated accordingly.
>  
> Using -R2 will obviously require more memory. Hopefully this will not
> be a big problem.
>  
> When done I will of course commit these new features/bugfixes to the
> cvs repository. And as I mentioned previously I'll be glad to recieve
> suggestions and comments, but please send them soon, for I'll go on
> vacation this sunday.
>  
> /Erik Marklund
> -- 
> Erik Marklund, PhD Student, Molecular Biopcysics group,
> Dept. of Cell and Molecular Biology, Uppsala University.
> Husargatan 3, Box 596,          75124 Uppsala, Sweden
> phone:  46 18 471 4537          fax: 46 18 511 755
> erikm at xray.bmc.uu.se
> 
> _______________________________________________
> gmx-developers mailing list
> gmx-developers at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-developers
> Please don't post (un)subscribe requests to the list. Use the 
> www interface or send it to gmx-developers-request at gromacs.org.
-- 
Erik Marklund, PhD Student, Molecular Biopcysics group,
Dept. of Cell and Molecular Biology, Uppsala University.
Husargatan 3, Box 596,          75124 Uppsala, Sweden
phone: +46 18 471 4537          fax: +46 18 511 755
erikm at xray.bmc.uu.se