[gmx-users] cross angle between two alpha helices

[Vincent Lemaitre]

> On Fri, 26 Apr 2002, YUJIE WU wrote:
> >Hi,
> >
> >I want to calculate the angle formed by two alpha helices. Does gromacs
> >have to a tool
> >for doing that? Thanks.
>
> try g_bundle, I don't know whether it is what you want though.

It is possible to do it with g_bundle, althought a bit dirty...
I have done it once like this:

# 1. Do in make_ndx:
# make_ndx	Generate indexes describing extremities of
# 		the two helices. Select the CA of residues belonging
#		the extremities of each helix.
make_ndx -f NAME.gro -o NAME_CROSSANGLE.ndx
# Define other groups you need
# Define 4 groups corresponding to the ends of the two helices
# ...
# q
	
# 2. Do in a text editor:
# Merge the 4 index groups, corresponding to helices ends, in order to get
# two index groups, each containing either the beginning of the two
# helices or the ends of the two helices. See g_bundle's online help.
# The number of residues selected for each helix has to be the same.
	
# 3. Do in g_bundle:
# g_bundle	Get coordinate of vectors describing the helices
echo 0 1 | g_bundle -f NAME.xtc -s NAME.tpr -n NAME_CROSSANGLE.ndx \
                    -dt XX -b YYYY -e ZZZZ -na 2 -oa NAME_CROSSANGLE.pdb\
                    >& g_bundle.CROSSANGLE.out		      
# -na 2		     	   ; there are two axes this time
# -oa NAME_CROSSANGLE.pdb  ; to get the coordinates of the axes and
			     check the angles given by the software

# rm, Remove some unuseful files genererated by g_bundle:
rm bun_len.xvg bun_dist.xvg bun_z.xvg bun_tilt.vxg bun_tiltl.xvg bun_tiltr.xvg
	
# The first time, it might be helpful to open NAME_CROSSANGLE.pdb in rasmol or
# vmd to visualize the data you are handling and check wether you define your
# index correctly... and also two understand the structures of the data you
# are handling.

# The next part is dirty and could be automated with a smart script:

# 4. Do in g_ traj
# g_traj transform the .pdb trajectory in a file format easier to handle,
# i.e a .xvg file.
echo 0 | g_traj -f NAME_CROSSANGLE.pdb -s NAME_CROSSANGLE.pdb \
		-ox NAME_CROSSANGLE.xvg >& g_traj.CROSSANGLE.out

#In the header of the .xvg file, you have this:
# @ legend string 0 "atom 1 X"
# @ legend string 1 "atom 1 Y"
# @ legend string 2 "atom 1 Z"
# @ legend string 3 "atom 2 X"
# @ legend string 4 "atom 2 Y"
# @ legend string 5 "atom 2 Z"
# @ legend string 6 "atom 3 X"
# @ legend string 7 "atom 3 Y"
# @ legend string 8 "atom 3 Z"
# @ legend string 9 "atom 4 X"
# @ legend string 10 "atom 4 Y"
# @ legend string 11 "atom 4 Z"
# @ legend string 12 "atom 5 X"
# @ legend string 13 "atom 5 Y"
# @ legend string 14 "atom 5 Z"
# @ legend string 15 "atom 6 X"
# @ legend string 16 "atom 6 Y"
# @ legend string 17 "atom 6 Z"
#  0	1.3543	1.2306	2.1121	0.8561	0.8589	...
#  20	1.4002	1.0692	2.0428	0.8943	0.7557	...
#  40	1.359	1.0793	2.1341	0.8686	0.7528	...
#  ...	...
	
# 5. Do in Microsoft Excel or equivalent
# Import the .xvg file. You'll get a 19-column file. The first column
# correspond to the time. The next 9 columns are (x,y,z) coordinates of 3
# points belonging to the vector describing the first helix and the next 9
# are coordinate of 3 points belonging to the second helix.
	
# When this point is reached, you can define the vectors describing the
# helices (by substracting coordinate of atom 3 and 1 and atom 6 and 4),
# normalize them (by dividing the vector by their norm) and calulate their
# dot-product in excel.
# The dot product of two normalized vectors is the cosine of their
# cross-angle.

Hope it can help...

Vincent


____________________________

Vincent Lemaitre
D.Phil Student

Biomembrane Unit
Biochemiistry Department
University of Oxford
South Parks Road
OX1 3QU Oxford
UK

Tel: +44 (0)1865 275 270
Fax: +44 (0)1865 275 234