Tuesday, 18 January 2011

Peter Murray-Rust, PPS and a Semantic Molecular Future

Yesterday (17 January) I attended a symposium at the Department of Chemistry, University of Cambridge, that was held to celebrate the career and ideas of one of the founders of the PPS course: Peter Murray-Rust. Since 2000, Peter has been based in Cambridge, where he is a Reader in Molecular Informatics.

The symposium was opened by Professor Sir Tom Blundell, a former head of the crystallography department at Birkbeck and now emeritus professor of biochemistry at Cambridge. Tom told the audience that when Peter came to Birkbeck in the mid-90s he already had a distinguished career in molecular science behind him. He had been a PhD student at Oxford in the 1960s, working with Keith Prout on the crystal structures of inorganic molecules. Tom, who worked with Dorothy Hodgkin on some of the early crystal structures of insulin, was a fellow student. Peter then worked as a lecturer at the University of Stirling, Scotland, and at pharma giant Glaxo (now part of GSK). He joined Birkbeck at a time when the Web was just beginning to open up the world of the Internet to the wider community. He, with David Moss, then a reader in crystallography at Birkbeck, and research associate Alan Mills, saw the potential for the web to extend the department's specialist teaching beyond the reach of those who could readily commute to central London, and PPS was born. The first, experimental course was delivered, free of charge, in 1995; within six years it would be incorporated into the full online-only MSc course.

Peter has never been one to let the grass grow under his feet. While he was busy creating and launching the PPS course, he was already thinking about how the still infant Web could be harnessed to allow data and information to be manipulated and understood rather than simply displayed. With Henry Rzepa from Imperial College, London, he developed Chemical Markup Language (CML) as a version of XML ("extensible markup language") for chemists. This is now very widely used. The first scientific paper written entirely in XML was published in 2001, although the editors of the journal concerned described it as "an interesting exercise, but [not] easy to deal with by any means". It is now being used for Microsoft's new chemistry add-in for Word: Chem4Word.

CML is described as a semantic language. The term "semantic web" was coined by WWW developer Tim Berners-Lee
as "a group of methods and technologies to allow machines to understand the meaning – or 'semantics' – of information". Peter's wide interests include, besides the automatic analysis of data in scientific publications - the development of virtual scientific communities, and he campaigns passionately for all scientific data to be freely available to all. He was one of a small group who drafted the Panton Principles (named after a Cambridge pub) which state that future scientific advances will depend crucially on all science data - not necessarily its interpretation - being made freely available on the Internet. Later presentations, including one by Henry Rzepa, developed these ideas in more detail. The symposium ended with a presentation of software being developed in Peter's group, including an application where a student manipulated an image of a molecule by waving his arms. This may have looked like a fun gimmick, but it must be potentially useful for disabled students who have difficulty with using a mouse.

There was one other delegate at the symposium to whom PPS students owe a debt, although they are probably unaware that they do: Roger Sayle, the developer of Rasmol. Today, we rely on molecular graphics programs that are fast, free, and easy to install and use on any desktop machine. Roger's Rasmol, developed in the mid-90s, was the first of these.

If you would like to know more about Peter and his ideas, and how some of them have been applied in Birkbeck's web courses, you can visit his blog.