Active Molecular Information and the ECTOC

Chemical MIME Types
Inserting Molecular Hyperlinks into ECTOC Papers
Thumbnail Images and Mapped Images
Configuring X-Mosaic Helpers
Configuring HTTP Servers
Configuring Macintosh Browsers
End of Line Characters
Configuring Microsoft Windows Browsers
Slow Network Response
Chemical Structure Markup Language
Saving and Printing
Further Information


The first Electronic Computational Chemistry Conference (ECCC) pioneered the use of the Internet and World-Wide Web for making available and discussing new research results in Computational Chemistry. The ECTOC conference is focusing on Topics in Organic Chemistry If you have quantitative molecular information available in the form of standard data files such as MOLFILE PDB, XYZ and other formats, then we strongly encourage you when writing the full version of your ECTOC paper to consider hyperlinking to e.g. co-ordinate files, output files and other supplemental information that you feel might enhance the value of your paper to others. Such "live" information is one of the exciting new possibilities that "electronic conferences" such as ECTOC can offer, which we hope will serve as a model for future chemical conferences.

Chemical MIME types are the currently favoured mechanism for delivering such information via the Internet. Although MIME can be used within the Gopher and Mail Internet protocols, it is best implemented within the World-Wide Web infra structure. This enables hyperlinks to be inserted into e.g. the text or image of a paper which points to molecular data stored in a standard format. When the hyperlink is activated, the molecular data will be automatically transferred to your computer and can be processed using a pre-defined "helper" program. Such "Internet modelling kits" have also been referred to as "hyperactive molecules".

These notes set out how to insert such molecular hyperlinks into HTML documents, how to acquire and specify molecular "helpers", and how to set up the mechanism on your own World-Wide Web server if you wish. The specific details have been organised according to the three principal types of computer most commonly used for WWW browsing. This document is NOT a HTML primer, on which you can get much more information here which contains a comprehensive FAQ list, or here which is the NCSA guide.

Back to index.

Inserting Molecular Hyperlinks into ECTOC Papers

Here is an example of a Microsoft Word document converted to its HTML equivalent using rtf2html, a program written by Chris Hector. This shows how a paper can be readily marked up. For a beginners primer, click here. A typical text link to e.g. a data file in say PDB format in such a document can be <A HREF="">accomplished thus</A>. The file name qualifier .pdb will be associated with the chemical/x-pdb MIME type by the HTTPD server configuration file on the server (in this case running as

The following list indicates the filename suffixes you will have to assign in order to add other types of document;

 .mol  MDL Molfiles
 .xyz  Files saved in the XMol format (highly concise and shorter than e.g. .gal)
 .pdb  The Brookhaven Format
You can use other formats, but you will have to define the appropriate MIME types.

If you want to use the ECTOC server to save your HTML and coordinate files, enter a <A HREF="1prc.pdb">local file reference</A> and send both the HTML file and the molecular data file to the ECTOC conference. Note also that local references can be globally converted to absolute references by the addition of a line of the following type to the header field at the top of the HTML document;
Whilst X-Mosaic appears to support this syntax, other browsers may not.

Thumbnail Images and Mapped Images.

If you want to draw attention to a molecular hyperlink, one effective way of doing so is to place it within a small thumbnail image of e.g. the molecule. This can be done as follows;
You can go one stage further and map various regions of a perhaps larger thumbnail image to different molecular hyperlinks, as follows;
You will also have to define a .map file containing the mapped regions of the image, and specify this map file in the imagemap.conf file located in /usr/local/etc/httpd/conf/ directory on your server. The best way of accomplishing this process is to inspect a previously completed imagemap installation.
Back to index.

Unix X-Windows Molecular Browsers

There are several free Molecular Visualisation Helpers available from ftp sites, including Babel is a Chemical File conversion utility which should help you interconvert various files formats into one of the ones supported by the above programs.

XMol is best suited for the visualisation of data derived from e.g. programs such as Gaussian 92, MOPAC etc, and can also be used for generating animations of reaction paths, IRCs, normal vibrational modes and dynamics runs. Installation is via a Makefile, althoug be warned that this generates quite a number of directories and files, and is best done as Superuser. A new release of XMol is expected shortly.

RasMol consists of just one single executable file which should be installed either in your own directory, or in a generic directory such as /usr/local/bin (as Superuser). and is the program of choice for the visualisation of large molecules. It also supports a highly flexible "markup" syntax, which can be used via a script such as CSML to annotate the structure in order to highlight particular aspects.

The Iris Explorer EyeChem suite offers the capability for display of a number of molecular properties, but currently only runs on SGI systems with Version 2.2 of Explorer.

Configuring X-Mosaic Molecular Helpers.

This is achieved via a so called .mailcap file located in your home directory. One or more of the following lines should be added, depending on what type of file you might wish to visualise. You must NOT define any MIME type twice, i.e. specify either Rasmol or Xmol for pdb files, but not both. Also included is the proposed Mathematica MIME type, for anyone wishing to include a Mathematica notebook to illustrate a theoretical point. For other entries, click here. If you know of a suitable free program that you think may be useful, or of well defined candidates for chemical MIME types, contact Henry Rzepa,
chemical/x-pdb;            rasmol %s 
chemical/x-xyz;            rasmol xyz %s 
chemical/x-pdb;            xmol -readFormat pdb %s 
chemical/x-mopac;          xmol -readFormat mopii %s 
chemical/x-mopac-out;      xmol -readFormat mopo %s 
chemical/x-gaussian;       xmol -readFormat gaussin %s 
chemical/x-gaussian-log;   xmol -readformat gausso %s 
chemical/x-xyz;            xmol -readFormat xyz %s
chemical/x-mopac-out;      eyemopac %s 
chemical/x-gaussian-log;   eyeg92 %s 
chemical/x-csml;           csml %s 
application/x-mathematica; math %s 
You must have a path defined to the relevant programs. If these are stored in other locations, you can specify the path as e.g.;
chemical/x-pdb;             /usr/local/bin/rasmol %s
There are also occasions when you may wish to open up a command line interface to the helper program (i.e. Rasmol uses this extensively);
chemical/x-pdb;            xterm -T title -e rasmol %s
Two windows will then appear on your screen rather than the usual one.

Configuring HTTPD Servers.

If you want to "serve" molecular information from your own World-Wide Web server, you will need to define suitable chemical MIME types. This is done by editing the file mime.types, normally found in the directory /usr/local/etc/httpd/conf. Add entries of the type;
chemical/x-pdb            pdb 
chemical/x-xyz            xyz
chemical/x-mopac          mop 
chemical/x-mopac-out      moo 
chemical/x-mopac-vib      mvb  
chemical/x-gaussian       gau 
chemical/x-gaussian-log   gal
chemical/x-csml           csml  
application/x-mathematica ma
After these modifications, you will have to kill the httpd process and restart it by typing /usr/local/etc/httpd/httpd
Back to index.

Macintosh Molecular Browser Helpers

On a Macintosh system the following programs are worth a look; Babel is a Chemical File conversion utility which should help you interconvert various files formats into one of the ones supported by the above programs.

Kinemage consists of two programs, a pre-processor to convert pdb files to "Kin" format, and the Kin viewer MAGE. "Kinemage" type files have been proposed as a chemical MIME type in their own right. Mail for further details.

Ball & Stick supports a wide variety of file types, including MOPAC, SHELX, MOLDAT etc, but the files have to be opened manually using the Import Menu item and are limited to the display of 40 atoms in the demo version.

Chem3D Viewer only reads files in CSC format, although the full version of the program can read MOPAC, PDB and a large variety of other files.

MicroWorld is an impressive display package, but does not read most file formats directly and requires that you use Babel to convert from the standard formats first.

RasMac for the Macintosh was released in October 1994. If you know of other freely distributable Macintosh Molecular visalisers, and in particular ones which do not require user intervention on startup (i.e. the ones above all throw up a user dialog before displaying an image), then Henry Rzepa would be delighted to hear from you.

Configuring Macintosh WWW Browsers.

The recommended browsers are MacWeb which appears stable at version 1.00alpha3, NCSA MacMosaic, of which only version 2.00A17 can be recommended as being reasonably stable and Netscape 1.0N. These browsers are quite similar in their configuration. A screen from MacWeb show the MIME definition. With Mosaic, select Apps from Preferences and define a new document type.

End of Line Characters

Because of the way the end-of-line character is defined on a Macintosh (CR) and Unix systems (LF), you may find that text files do not translate correctly after transfer from a Unix based httpd server. This would only occur when the MIME type is defined as other then text, as in the chemical examples shown above. The MacWeb browser enables LF => CR conversion to be specified for any MIME type, not just for text files.

To use MacWeb, add the following MIME types to the configuration;

 chemical/x-mopac          (suffix mapping .mop, 7-bit EOL translations) 
 chemical/x-pdb            (suffix mapping .pdb, 7-bit EOL translations) 
 chemical/x-kinemage       (suffix mapping .kin, 7-bit EOL translations) 
 application/x-mathematica (suffix mapping .ma,  7-bit EOL translations)

Back to index.

Microsoft Windows

MS Windows versions of both Kinemage and Rasmol are available. Other demo versions of programs such as DTMM, Nemesis etc are available. If you know of other freely distributable Molecular visalisers running under Windows (or Windows NT), and in particular ones which do not require user intervention on startup, then Henry Rzepa would be delighted to hear from you.

Configuring Microsoft Windows Browsers.

A Sample Mosaic.ini file in which the MIME types and Helpers are defined is available here. This should be placed in the Windows directory. WinWeb will require a similar WinWeb.ini configuration.
Back to index.

Delaying Image Loads

If you are experiencing difficulties with network bandwidth in downloading some of the larger images, you may wish to delay image loading. On X-Mosaic, add the following line to the Mosaic configuration file stored as /usr/lib/X11/app-defaults/Mosaic
Mosaic*delayImageLoads: true
On Mosaic for Windows, a similar line has to be inserted into the Mosaic.ini file. On Macintosh NCSA Mosaic 2.0 and MacWeb this can be achieved via the preferences menu items.
Back to index.

Chemical Structure Markup Language

The mechanisms described above enable the visualisation of molecular data using local programs, via hyperlinks to WWW based material. Often however, it is desirable to annotate on-the-fly the molecular structure in the context of specific discussion in a hypertext document. Whilst some commercial modelling packages enable such annotation scripts to be invoked (the .spl scripts in Sybyl for example), these tend to be proprietary in nature, and also have no facile mechanism for communicating with a WWW browser.

We have implemented a simple such mechanism for annotating a Rasmol display, which we have found particularly useful for enhancing navigation through a complex biomolecule via the WWW mechanism.

To implement this mechanism on an X-Window workstation, you will need to first acquire Tcl/Tk if it is not already installed on your system (most do nowadays). In the spirit of the html markup language, we have defined a simple chemical structure markup language which is interpreted by a short script resident on the local machine. The way it all works is as follows. Ensure that your .mailcap file has an entry of the following type
chemical/x-csml csml
If you are running a HTTPD server, the mime.types file has an entry of the type;
chemical/x-csml csml
Place the CSML script in a suitable directory such as /usr/local/bin. A typical set of CSML instructions might take the form;

gly39 red spacefill
atomno=410 blue dot
chain yellow ribbon
and would be saved on the HTTP server with the suffix .csml in the filename. A hyperlink is then made to this file from a suitable point in a document. When activated, the file is copied to the local system, and run as the input to the CSML script. This script ensures that the instructions contained in the file are passed to a running Rasmol process which ensures that the selected residues or atoms are suitably highlighted. To see an example of this process in action, click here.
Back to index.

Saving and Printing

Downloading molecular information by activating a hyperlink allows near instant visualisation of the information. However, this information is stored by the WWW browsers in temporary files which do not normally persist. The option to save the files permanently is normally an obvious selectable option within the browser. MacWeb implements a rather neat solution, which is to press the shift key whilst activating the hyperlink. Mosaic implements a "save to disk" switch which however applies globally to all downloaded files.

If you wish to print a document, all WWW browsers support printing. However, the results can be highly variable. X-Mosaic supports postscript printing, but appears to handle the color map of in-lined images rather oddly. Mosaic for Windows does rather a good job of rescaling in-lined images, although beware that page throws may not be handled particularly intelligently.

Back to index.

Further Information.

Last Updated 18 January 1995. If your have any comments or questions about this document, contact Henry Rzepa; A forthcoming book from ACS publications edited by Steven Bachrach will contain chapters on all aspects of the use of the Internet in Chemistry.