[gmx-users] Re: gmx-users Digest, Vol 68, Issue 35

[Mark Abraham]

Stefan Hoorman wrote:
> 
> 
>     Stefan Hoorman wrote:
>      >
>      >     Stefan Hoorman wrote:
>      >      > How can I calculate the angle between a helix inserted in a
>      >     membrane and
>      >      > the axis perpendicular to the surface of the membrane. I
>     have tried
>      >      > using g_helixorient, but the graphs all come as a straight
>     line
>      >     in zero.
>      >
>      >     See -z in g_sgangle -h
>      >
>      >     Mark
>      >
>      >
>      > Thank you for the tip, but again I faced a problem. I could only
>     analyse
>      > using g_sgangle with an index group containing 3 atoms. Then the
>     problem
>      > is, since my helix not only tilts, but also it bends a little
>     back and
>      > forth, the angle between it and the z axis will vary greatly
>     depending
>      > on which three atoms I choose. I imagined that by choosing three
>     alpha
>      > carbon atoms every four residues would give me a good outcome, but it
>      > turns out that the angle changes a lot, and I am guessing it is
>     because
>      > of this bending movement. When I say "changes a lot" I mean; it
>     starts
>      > at 0º and gets up to 50º, which i imagine to be a little bit too
>     much.
>      > Would there be a better way to calculate this?
> 
>     g_principal will calculate the axes of inertia, from which you can
>     calculate the angle with the Z-axis.
> 
>     Mark
> 
> Ok, thank you for the info, but now I got a little bit confused. What 
> exactly does the info from axis1.dat, axis2.dat, axis3.dat and moi.dat 
> mean?

I don't know, I've never seen them. I'd be expecting three numbers in 
each to indicate a vector parallel to the axis.

> And how would I use this info to calculate the helix tilt. I know 
> the general meaning of principal axes of inertia, but I could not find 
> any information more specific to this type of caltulation I need to do. 
> If you could give any reference or any other type of information a 
> little bit more specific I would be very greatful. I mean, if it is not 
> too much to ask.

This is standard geometry - dot-products of vectors are related to the 
cosine of the angle between them. Check wikipedia.

Mark