Hello, PyMOL users,

The driving force behind these last four months of Jmol/JSmol development has been e-publication. The goal has been to produce a means of e-publishing working interactive biomolecular models that are ubiquitously accessible (i.e. no Java, no  WebGL). We achieved that goal for Jmol late last year, with the introduction of "JSmol" -- a fully HTML5-based identical renderer for Jmol. See http://chemapps.stolaf.edu/jmol/jsmol/jsmol.htm

Today, I'm letting you all know that we have that for PyMOL as well, basically. Authors can produce all the beautiful PyMOL images they want for regular publication purposes and then have them available for general viewing on a publisher's web site (if that site is so equipped) or on their own institutional site.

This works because Jmol now has a pretty decent PyMOL session file reader. It's a work in progress, but we are reproducing our complete set of test files (104 to date) fully now, with quite a wide variety of features. It's a work in progress, of course, and we are handling issues as they arise. But I think one look at http://ispcsrv3.weizmann.ac.il/a2jmolb/browse will convince you that we have made some pretty significant progress along these lines.

[That server is having significant issues just this week, so pardon the "Failed" signs. Those aren't real problems; just some sort of unidentified server issue at the Weizmann Institute.]

There are some differences in that test set, but they are only cosmetic -- a few sightly different renderings of cartoons, for instance. Jmol, though it has a PovRAY interface, doesn't interactively do shadows. Translucency is handled differently. (PyMOL's translucency is object-based; Jmol's translucency is pixel-based.) The perspective is exact, though, because both programs use the same perspective model. Morph-type (not scripted) movies are no problem. Electron density maps are read. Scenes are cataloged, reproducible, and easily "discovered" by web page based JavaScript.

Of course, the JavaScript-only non-WebGL version is slower than Java, which itself is (sometimes perceptively) slower than C++. That's a given. Don't expect prime performance on your smart phone.

Surface generation, particularly the PyMOL default solvent-excluded surface, is considerably slower in JavaScript, but we have a way around that. There's a way of caching the session files in Jmol that allows them to be displayed essentially instantly, even in JavaScript.

This will be released as Jmol 13.2.0 sometime next week. In the mean time, if you want to experiment --- realizing that most of you do not do anything with Jmol right now, right? -- that's understandable --- the link to the compiled Java and HTML5 versions are:

(Java, including stand-alone Jmol.jar and JmolData.jar and applet files)

(HTML5/JavaScript and also applet files, with demo pages)

I'm hoping this will interest some of you, and you might get involved in future efforts as we develop this connection further.

Bob Hanson

Robert M. Hanson
Larson-Anderson Professor of Chemistry
Chair, Chemistry Department
St. Olaf College
Northfield, MN

If nature does not answer first what we want,
it is better to take what answer we get.

-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900