Here you'll find a small gallery section of pictures and movies I did mostly last year for several people of my previous lab (Strasbourg, France). The structures used have all been solved there.

Protein and ligand objects (ribbons, coils, tubes, and so on) were generated seperately by a modified version of Molscript-2^web  which provides POV-Ray output. This is another patch I wrote a long time ago; it is not available here for several reasons, including the following. You can obtain similar 3D object using other softwares that output structures in POV-Ray format. See the Links section for a few examples.

Playing the animations below requires software decoders that handle MPEG-1 and/or MPEG-2 program streams (basically DVD players). There are lots of players for Windows and probably a few for Macintosh. Unix/Linux users might have a look at the following suggested ressources:

1. The excellent MPlayer^web  (recommended)
2. The VideoLan project^web  at Centrale Paris
3. The Xine player^web  at SourceForge
4. The mpeg2dec and OMS/OMI players at LiViD^web 

Click to enlarge (1024x768)

This picture was not created in Pov4Grasp but rather using the official version of POV-Ray 3.50cweb  and grasp2pov. The goal was essentially to test the radiosity improvements over POV-Ray 3.1g (on which Pov4Grasp is currently based). Radiosity is also called "global illumination": all objects influence the lighting of each other via inter-diffuse reflections. Here, the most visible effect of using this expensive technique is due to the white floor. Without radiosity, all parts of the protein and the surface not directly lit by the main light source would be completely dark, as in the pocket where the red ligand is bound. The surface was created by MSMSweb  and converted with Lothar Esser's msms2srfweb utility. It was then imported into GRASP to calculate and map the electrostatic potentials on the surface. The new surface file written by GRASP was finally converted to a mesh2 object using the -megapov option of grasp2pov and updating the #version directive in the output file. The image was actually rendered at 1800x1350 for printing quality.

Click to enlarge (800x800)

Oestrogen Receptor dimer. These are two different surfaces, each of them using a plain color and clipped by a manually-adjusted box which englobes only half of the surface.

Context. This animation is the second I prepared to present the structure of Stromelysin-3 (another very simple and short movie is found on this page). Its purpose is mainly to show that this structure exhibits a tunnel filled with an hydrophobic sidechain of its bound ligand and two water molecules. The animation has been presented by Anne-Laure Gall at the Paul Basset Memorial meeting (IGBMC) in September 2000 and later on at her PhD Thesis.

Technics. Camera if flying toward and inside the molecular surface. A high-detailed object (more than 500,000 triangles) was required when travelling through the tunnel and especially when leaving it. The camera follows two spline paths: one for its space location, and another for its look_at vector. Its speed is also spline-controlled in order to get smooth acceleration/deceleration stages. Camera "records" motion blur -as designed in Pov4Grasp- resulting in a simulated shutter speed of 1/50 s (see snapshots below). Preparation was done by hand, with a lot of rendering tests using the "preview" flag in the grasp_surface object definition. This took roughly one month of precise tweaking and adjustements.

Rendering. Using Pov4Grasp's "persistent" grasp_surface object, the rendering of 750 frames lasted approximately 50 CPU hours on a DEC alpha EV6 with 512 Mb of on-board memory. Predicted time with official POV-Ray 3.1g for similar result (without motion blur) is about 4 months, mostly spent when parsing surface data. Actually this is under-estimated since our workstation was unable to render a single frame due to memory requirements. In other words: this animation could not be calculated without Pov4Grasp.

		Anamorphic frame (720x576) Displayed image (1024x576)
		Anamorphic frame (720x576) Displayed image (1024x576)
		Anamorphic frame (720x576) Displayed image (1024x576)
		Anamorphic frame (720x576) Displayed image (1024x576)
		Anamorphic frame (720x576) Displayed image (1024x576)
		Anamorphic frame (720x576) Displayed image (1024x576)

Animation (30 seconds) is available in two formats:

1. MPEG-1 stream (4.2 Mbytes) in PAL VCD format (352x288 @ 25 fps, no sound).
2. MPEG-2 stream (18.4 Mbytes) in PAL DVD format (720x576 @ 25 fps, no sound).
   This is an anamorphic widescreen presentation (16:9 ratio) so you may experience image distorsion if your player does not scale it properly. The movie originally contained a Dolby Digital 2.0 audio excerpt from the score of Ridley Scott's film "Blade Runner". This music by Vangelis is not included here due to Copyright restrictions.

This second animation is showing the Vitamin D molecule being docked into the pocket of its nuclear receptor. The docking is actually fake: it was not calculated at all. Once the ligand is properly fitted, the pocket is closed by the H12 helix in its agonist conformation. The camera uses heavy motion blur for the first part (see first snapshot). This animation was both prepared and rendered in one night (well, pretty short deadline !).

		Anamorphic frame (720x576) Displayed image (1024x576)
		Anamorphic frame (720x576) Displayed image (1024x576)
		Anamorphic frame (720x576) Displayed image (1024x576)
		Anamorphic frame (720x576) Displayed image (1024x576)
		Anamorphic frame (720x576) Displayed image (1024x576)

Animation (30 seconds) is available in two formats:

1. MPEG-1 stream (4.2 Mbytes) in PAL VCD format (352x288 @ 25 fps, no sound).
2. MPEG-2 stream (11.9 Mbytes) in PAL DVD format (720x576 @ 25 fps, no sound).
   This is another anamorphic widescreen 16:9 presentation.

Page last updated: Apr 17, 2007 Any comment or suggestion ?