Gorgeous method

Gorgeous

Run RasMol, and rotate the model (tcgs) to the position you want. (on Mac)

(Nice RasMol tutorial is presented by Dr.Okamoto, Osaka Med. Univ.)

Write molscript files from ' RasMol Command Line'. (on Mac)

    
    write molscript mol.1
    
    rotate y -6
    
    write molscript mol.2

These commands make files on RasMol directory, name 'mol.1' and 'mol.2'.

Content of files...

! File: mol.1
! Creator: RasMol Version 2.7.2.1
! Version: MolScript v1.3
  ...
  transform atom *
    by centre position atom *
    by rotation x 180.0
    by rotation x 79.7272
    by rotation y 4.32451
    by rotation z 123.309;

... omit

! File: mol.2
! Creator: RasMol Version 2.7.2.1
! Version: MolScript v1.3
  ...
  transform atom *
    by centre position atom *
    by rotation x 180.0
    by rotation x 74.6711
    by rotation y 7.59851
    by rotation z 122.783;

... omit

Only rotation angles are differents. These angles are neccesarry!

Run molscript. (on Linux)

$ molscript < mol.1 > del.ps
----- MolScript v2.1.2, Copyright (C) ...
----- ref: P.J. Kraulis, J. Appl. Cryst ...
----- http://www.avatar.se/molscript/
reading PDB file...
13 residues and 220 atoms read into mol...
220 atoms selected for position
220 atoms selected for transform
rotation matrix applied:
-0.54759064 0.18978375 -0.81493964
0.83334179 0.03592585 -0.55158934
-0.07540530 -0.98116842 -0.17782739
translation vector applied:
15.3691 -2.3984 11.8403
coil from 1 to 10 (10 residues)
...

$ molscript < mol.2 > del.ps
----- MolScript v2.1.2, Copyright (C) ...
----- ref: P.J. Kraulis, J. Appl. Cryst. ...
----- http://www.avatar.se/molscript/
reading PDB file...
13 residues and 220 atoms read into mol...
220 atoms selected for position
220 atoms selected for transform
rotation matrix applied:
-0.53670422 0.29130457 -0.79189028
0.83334485 0.03592487 -0.55158477
-0.13223061 -0.95595557 -0.26203819
translation vector applied:
14.0473 -2.3985 13.3820
coil from 1 to 10 (10 residues)
...

Keep/save these real rotation matrix. NCSA telnet is nice for cut & paste on Mac window.

Edit anisotropic PDB and reduce B values of hydrogens. (on Mac)

Generaly H-atoms have fairly big thermal factors, and they are nuisance in plotting. Therefore, B values of H-atoms are changed to uniformed 0.9 here.

ATOM      8  HG1 VAL     1      15.999   4.970  13.598 1.000  9.70
ATOM      9  HG1 VAL     1      15.229   3.758  14.225 1.000  9.70
ATOM     10  HG1 VAL     1      16.077   4.706  15.141 1.000  9.70
                            |
                            V
ATOM      8  HG1 VAL     1      15.999   4.970  13.598 1.000  0.90
ATOM      9  HG1 VAL     1      15.229   3.758  14.225 1.000  0.90
ATOM     10  HG1 VAL     1      16.077   4.706  15.141 1.000  0.90

Changed data: tcgs2.pdb.

Run Raster3D/rastep for anisotropic PDB. (on Linux)
```
$ rastep -auto -fancy3 -radius 0.12 -prob 0.4 < tcgs2.pdb > tcgs2.r3d
```
'-auto' determines proper rotation matrix
'-fancy3' selects ellipsoid types, No.1 and 3 are transparent surface (see manual)
'-radius' defines bond radius
'-prob' defines probability of ellipsoid size

Make two copies of tcgs2.r3d and change rotation matrix. (on Linux)

$ cp tcgs2.r3d tcgs2L.r3d
$ cp tcgs2.r3d tcgs2R.r3d

These file contains rotation matrix calculated automatically by rastep...

Raster3D thermal ellipsoid program V2.6e      40% probability bounds
80  64    tiles in x,y
 8   8    pixels (x,y) per tile
4         3x3 virtual pixels -> 2x2 pixels
1 1 1     white background
F         no, shadows are dorky
25        Phong power
0.15      secondary light contribution
0.05      ambient light contribution
0.25      specular reflection component
   ...
# Auto-orientation matrix
16
ROTATION
      0.55373     -0.57131      0.60580
      0.79441      0.14438     -0.58997
      0.24959      0.80794      0.53380
# End auto-orientation
   ...

Change only rotation matrix of tcgs2L.r3d for mol.1, and tcgs2R.r3d for mol.2
(easy! edit Linux file from NCSA telnet/Mac using cut&paste).

Changed data: tcgs2L.r3d and tcgs2R.r3d

Run msms to make ASA data. (on Linux)
```
$ cat tcgs_nonsolv.pdb | pdb_to_xyzr > tcgs.xyzr                       # conver format
$ msms -if tcgs.xyzr -of tcgs -probe_radius 1.4 -density 5.0 
```
msms produces two files, 'tcgs.face' and 'tcgs.vert'.
Keypoint here: ASA is calculated for all atoms in PDB file. I remove the solvent from PDB ('tcgs_nonsolv.pdb'), and ASA was calculated for only peptide.

Run Raster3D/msms2r3d to convert suface date to raster3D. (on Linux)

Raster3D package (2.1.2e) offers a converter from msms (VRML) to Raster3D, and also examples. They are located in '~/Raster3D_2.6e/examples/msms'.
Convert is...

./msms2r3d.pl tcgs.face tcgs.vert > tcgs.surf.r3d

Now, we have everythig to draw ASA-overlayed elliposid stereo view!

Check things we have...

tcgs2L.r3d tcgs2R.r3d	ellipsoid data for left and right eyes (rastep made)
tcgs.surf.r3d	converted surface data (msms made)
transparent.r3d	transpancy factors (located at ./examples). see manual! transpancy level is controled by this file.

Run Raster3D/render. (on Linux)

$ render -size 800x800 -zoom 85% -png tcgs_L.png -shadow << eof
  @tcgs2L.r3d
  @transparent.r3d
  @tcgs.surf.r3d
  eof
$
$ render -size 800x800 -zoom 85% -png tcgs_R.png -shadow << eof
  @tcgs2R.r3d
  @transparent.r3d
  @tcgs.surf.r3d
  eof

Caution for the order of files: 'transparent.r3d' works for following file.

Result...

Combine tcgs_L.png and tcgs_R.png to side-by-side by using graphic tools.

Can you see stereo?

Bigger one is GIF (248KB), or JPEG (144KB).
Oringal PNG files (800x800) are 224KB each.

More trials...

Surface color.

Surface data 'tcgs.surf.r3d' is like this...

1
5.063 4.645 18.918 4.711 4.869 19.218 4.618 4.272 18.675 0.866 0.866 0.866
7
0.100 -0.844 0.528 -0.076 -0.732 0.678 0.418 -0.577 0.701
1
4.618 4.272 18.675 4.711 4.869 19.218 4.266 4.496 18.975 0.866 0.866 0.866
...

I think the last three numbers are uniformed in the file and they represent Red, Green and Blue levels (0-1 each). The value '0.866 0.866 0.866' means relatively whity GRAY. The substitution of these result in colored surface.

These values are changed to '1.0 0.0 0.0' ('surf.red.r3d').

1
5.063 4.645 18.918 4.711 4.869 19.218 4.618 4.272 18.675 1.000 0.000 0.000
7
0.100 -0.844 0.528 -0.076 -0.732 0.678 0.418 -0.577 0.701
1
4.618 4.272 18.675 4.711 4.869 19.218 4.266 4.496 18.975 1.000 0.000 0.000
...

And then, run rastep.

$ render -size 800x800 -zoom 85% -png tcgs_red.png -shadow << eof
@tcgs2L.r3d
@transparent.r3d
@surf.red.r3d
eof
$

Woh, it's samewhate vulgar. Red surface was bad idea. Bigger one is here (JPEG, 56KB)

Make gel-like surface using contrast for background.

File 'tcgs2L.r3d' defines background color.
```
80  64    tiles in x,y
 8   8    pixels (x,y) per tile
4         3x3 virtual pixels -> 2x2 pixels
1 1 1     white background
...
```
Change '1 1 1' to '0 0 0' ('tcgs2Lb.r3d'). This means RGB levels are all zero (black).

File 'transparent.r3d' defines tranpancy level of object.
```
9  Here is an example of specifying a transparent material
8
  15.  0.6  -0.5 -0.5  0.5   0.80  0 0 0 0
```
The value '0.8' is changed to '1.0' ('transparent.100'). This means perfect invisible surface, and only reflected light are visible. Furhtermore, black background showups faint surface.

And then, run render.
```
$ render -size 800x800 -zoom 85% -png tcgs_b100.png -shadow << eof
@tcgs2Lb.r3d
@transparent.100
@tcgs.surf.r3d
eof
$
```
It's like as gel-encapsuled molecule.
Bigger one is here (JPEG, 56KB).
Make skelton surface using contrast for background.

Surface data 'tcgs.surf.r3d' is changed to 'surf.black20.r3d'. Surface color is little bit dark.
```
1
5.063 4.645 18.918 4.711 4.869 19.218 4.618 4.272 18.675 0.20 0.20 0.20
...
```
And then, run dender.
```
render -size 800x800 -zoom 85% -png tcgs_b100b.png -shadow << eof
@tcgs2Lb.r3d
@transparent.100
@surf.black20.r3d
eof
$
```
It seems to be molecule being in babble.
Bigger one is here (JPEG, 56KB).
Rule!: When the surface color is close to the background color, the appearance of surface became being transparent (thin film).

Do you like?

Back