[gmx-users] Re: g_chi again

[Graham Smith]

> the i j k l  in chi.log do not correspond to atoms describing phi,
>psi and omega. Does anyone have an explanation for that? What could I
> have done wrong?

You're not doing anything wrong - it's the program, which uses
non-standard definitions including the amino H. It's fine for chis
(especially with my modifications :-) ) but for backbone dihedrals you
might be better off using g_rama. There's also a program called
"chandran" on the contributions page that I think does futher
analysis.
BTW I got a mail last month from someone called Hender Grievink
<grievink at vvtp dot tudelft dot nl> asking about the same thing, 
so if you were thinking of trying to fix the g_chi code then I 
should mail him first in case he's working on the same thing. 

Graham