the Protein Picture Generator

This service is aimed at providing an easy way to generate pictures of protein structures, with the concern of integrating the most frequently used concepts of the molecular graphics fields. The design of the interface results from discussions with potential users about the features they would like to be implemented. The underlying machinery acts like a black box, but some understanding about the features of the interface might come from the knowledge that we presently use the  Dino program to produce the images, coupled with external programs such as stride (secondary structure determination), hbplus (hydrogen bond determination) or msms (molecular surface calculation).

- June-August 2004: First implementation by Cedric Binsiti and Salim Ahmed.
- September 2004: First development release of the  service, restrained access. Many improvements, in particular, users express the need for animations.
- Late October 2004: Second development release. Access set free to world, information to some Paris teams.
- December 2004: Pre-production release. Minor bug fixes. Server can treat several simultaneous requests.
- January 2005: Full production release, version 1.0.
- March 2005: Bug fixes (DNA), orientation previewer based on JMol, version 1.1.
- December 2005: Version 1.2, new previewer (sliders to reach interactive pre-orientation)
- June 2006: Version 1.21, small bug fix for side chains management in the default representation.

1. Design of the form
2. Data specification
3. Default representation
4. Supplementary representations
5. Process
6. Advanced parameters
7. Citations
8. Feedback
9. Gallery / examples

1. Design of the form:

The form is divided in four sections concerning respectively:

2. Data specification:

The data to render must be in the Protein Data Bank format (PDB).
It can be specified either on the form or a PDB Id , or by uploading the data. PDB Ids must be on the form: 1tim for the whole 1tim PDB entry, or 1timA for the chain A of the 1 tim entry. PDB identifier (the first 4 characters) are not case sensitive, hence 1Tim, 1TIM or 1tim are equivalent.
Uploading data or specifying a PDBId, you can also specify a list of chain Ids in the "Chains to visualize" field (e.g. AB, or A,B to select chains A and B of the PDB entry). Specifying "All" or nothing  will result in keeping all chains. Blank is a valid chain identifier.  Chain identifier are case sensitive. 1timA and 1tima are NOT equivalent.
Currently, it is possible to render only one PDB file. We plan to improve this soon. To make an image including several PDB entries, you can however concatenate them into one file, labeling each file using separate chain Ids. 

3.  Default representation:

3.1 Title: This character string eventually specified here will be automatically inserted into the picture. Multiple line text are possible using "\n" line break convention. Advanced parameters allow the customization of its aspect (font family, size and face) and position.

3.2 Display: This is to specify what will be displayed from the whole molecule.

For each of this topics, you can choose a specific rendering mode, and coloring pattern. I hope their names are rather self explicit.
Note: For each of the backbone, side-chains, ... surface sections, you can choose None, in order to adjust the representation using the supplementary representations
Coloring pattern notes:
Drawing modes:
ball and sticks                                       spheres                                             trace                                  

cartoon                                                      lines                                          spline

Colouring patterns:
The colours associated with each of these patterns can be modified in the advanced parameters section.

3.3 Scene parameters.

These parameters specify the parameters of the scene. The default values will ensure that the protein is at the centre of the image. You can however specify precisely the orientation of the protein, by filling the fields "Centre on" and "Focus on". This will specify a line that goes from the point specified as the centre to the eye of the user, passing through the point defined as the focus. Additional rotations can be applied (see the section Optional parameters).
3.4 Adjusting structure orientation

This set of parameters allow to adjust the orientation of the structure by small rotations (degrees).
The axes are defined as follows:
axes definition
The X and Y axes are within the plane of the drawing window, the Z axis is perpendicular to it. The arrows indicate the positive rotations.
For a set of values, the rotations are applied to the structure sequentially, following the order specified by the "order" option. Note for a same set of values the resulting orientation will differ for different rotation orders.
Since it can be difficult to assess the values to obtain some desired orientation, a special facility designed for that purpose can be accessed using the "preview orientation" button. Once a satisfactory set of values has been identified, you can simply report them from this help facility. Note that if you have specified a "Focus" (see previous section), the rotations apply after the focus transformation has been performed. If you modify the focus, the rotation values will become meaningless.

3.4 Picture parameters.

4. Supplementary representations:

This facility provides a mean of rendering some subparts of the molecule using a representation different from the default representation. Up to four independent selections can be specified.
For each, you need to specify the subset, its rendering mode, and its coloring pattern.
Note: For each selection, specifying Display as: None will result in applying the colouring pattern on the default representation.

Presently, although this might evolve, selections correspond to close to Dino valid selection expression, but are restricted, in particular concerning keywords (see below).

Example to select the backbone of TYR and ARG within residues 25 to 58 and 70 to 92:
rname=TYR,ARG and rnum=25:58,70:92 and not aname=CA,N,C,O
Complex selections such as a selection of the form :
(rname=TYR,ARG and rnum=25:58,70:92) or (rnum=25:58,70:92 and aname=CA,N,C,O)

i.e. a group of 2 elementary selections is also valid. Here, this selects all TYR and ARG for residue numbers 25 to 58 and 70 to 92
and atoms CA,N,C,O (heavy atoms of the backbone) for residues
25 to 58 and 70 to 92.

Valid keywords:
protein, dna, rna
backbone, solvent, hydrophobic, basic, acidic, polar, aromatic, aliphatic
purin, pyrimidin, base

backbone and rnum=25:58 and aromatic
Note: keywords are NOT used with "$".

Important Note: to avoid confusion, the "backbone" and "base" keywords correspond respectively to the protein backbone atoms and the nucleic acids base atoms. Hence, "protein and base" or "dna and backbone" lead to void selections.
Use: "(protein and backbone) or (dna and not base)" for a selection of the protein and nucleic acid backbone.
"(protein and not backbone) or (dna and base)" for a selection of protein side chains and nucleic acid bases.

5. Process:

This will launch the computation.
Depending on the complexity of the request and the server load, it might take from seconds to several minutes.
The results will present the pictures (for the formats supported by you browser), and a version of the script used to render, but not preserving the file names, since these are necessarily different on you computer and on the web server.  No acces to the MSMS  files describing the molecular surface (if necessary) is provided. You need to install
MSMS on your side to produce these.

6. Advanced parameters:

This section allows the customization of some of the most important parameters. It is organized by groups of parameters.

7. Citations:

When publishing images using surface rendering, please cite:
Sanner, M. F., Olson A. J. & Spehner, J.- C. (1996). "Reduced Surface: An Efficient Way to Compute Molecular Surfaces." Biopolymers 38: 305- 320.
HBonds calculation:
I.K. McDonald and J.M. Thornton (1994), "Satisfying Hydrogen Bonding Potential in Proteins", JMB 238:777-793.
Secondary structure identification using stride:
Frishman D, Argos P." Knowledge-based protein secondary structure assignment." Proteins. 1995 Dec;23(4):566-79.
DINO: Visualizing Structural Biology (2003)
Ansgar Philippsen

8. Feedback:

Please send comments, suggestions, images to add to the gallery to: tuffery's email

9. Gallery / examples:

PDB entry 1ggm
default view

PDB entry 2acy: 3 orthogonal views.
Default representation supplemented by aromatic residues
coloured by residue type


Example of the trypsin active site. PDB entry is 3tgi.
Backbone default representation is set to "None". The focus is on I.LYS15. The view angle is 30. An additional rotation of 30° on the Y axis is performed. The title text is:
"Trypsin (3tgi)\nDetail of the active site\nH57,D102,S195"
Supplementary representations:
1: "chain=E" display as cartoon, color using secondary structure
2: "chain=E and rnum=57,102,195 and not backbone" color according atomic types
"chain=E and rnum=57,102,195 and backbone" color as "ivory"
The background is set to grey, text color is lemonchiffon
Secondary structure colors are green for strands, yellow for helices.


(PDB entry 3cro).

Default parameters. View angle = 30

Example of combining only supplementary representations on the 1eyu PDB entry.
All default representations set to None. Rotation on X by -30 degrees.
Supplementary representation 1 set to: protein, displayed as cartoon, coloured by secondary structure.
Supplementary representation 2 set to: dna and not base, displayed as spline, couloured by chain (chain D - 4th chain - using gold).
Supplementary representation 3 set to: dna and base, displayed as balls and sticks, couloured by residue type.


PDB entry 1art, surface with opacity 0.75 coloured  according to temperature factors
The default backbone representation is preserved.
Heteros groups are displayed as spheres, coloured by atomic types.
View angle set to 30, additional rotation around X by 30° .


Cytosine deaminase. PDB entry 1rak.
The solvent is displayed around the molecular surface, transparent to show the backbone and the hydrogen bonds.
Optional parameters: Surface, and solvent set to all.
Image format: gif, animation: rock
All other parameters to their default values.

Mechano sensitive channel.
The view is on the axis of the channel.
Centre is set to: 27.498 128.966 7.763e-06 (centre of mass)
Focus is on: 25.966 127.680 8.459e-06 (displacement along the principal inertia axis of the structure)
The surface is coloured according to Temperature factors
Gif format. The animation is a translation along Z
Thanks to C. Etchebest and F. Guyon.


Comparing two sets of aromatic side chains conformations.
No defaul display. The input file was generated by concatenating two PDB files, assigning each a chain label (A and B).
selection1 set to "chain = A and not backbone and (aromatic)", coloured by residue type
selection2 set to "chain = B and not backbone and (aromatic) ", coloured as grey