Date: Mon, 03 Apr 1995 From: Sayle Dr R A Subject: Re: Suggestions To: Eric Martz You also asked for a short history of RasMol. I did my first degree at Imperial Collge, Univeristy of London. I spent my first summer working there working in the Computer Graphics section preparing images for a text book they were writing on the subject, published by Addison Wesley. Because of this work, i.e. I'd written the questions in the course text, I was advised not to do computer graphics as one of my final year options, but instead choose to work on parallel and distributed programming. It was about this time, 1989, that I met some crystallographers, in the Biophysics Department at Imperial. Their original "vector" based Evans & Sutherland PS300 displays were being replaced with new Silcon Graphics workstations. Biochemistry was shifting its emphesis from bonds and connectivity to surfaces and interactions. This resulted in a loss of depth information from the depth-cued (hither-and-yon) that was available with vectors and dots. The two possible ways of increasing depth perception was to add shadowing from a light source and allow images to be manipulated in real-time. At that time these were mutually exclusive, some of the images that I'd generated two years earlier had taken 9 hours to generate with shadows. This seemed to perfect final year parallel programming project. Spheres were the easiest graphical object to draw, and hopefully the parallel machines available at the time would achieve some speed-up on the problem. Hence the birth of RasMol, for shadowed RASter MOLecules (the fact that RAS are my initials was probably conincidental :). My final year project "Parallel Algorithms for Molecular Graphics" was then spent comparing different algorithms for speeding up ray-tracing on parallel computers. When I finished I'd won the University of London prize and had the second fastest program in the world for displaying shadows on spacefilling molecules. The fastest at this time was an implementation of Raster3D running on the 800 processor Edinburgh Parallel Supercomputer. For completely, different reasons I did my PhD in the Computer Science department at the University of Edinburgh, studying the ways that programs could automatically be made more efficient. While there I met Dr Andrew Coulson, the mentor of RasMol, in the Biocomputing Research Unit. I discovered that the improvements that made Raster3D go faster could also be applied to RasMol. Slowly, in my spare time, RasMol with and without shadows, got faster and faster on a parallel computer untill the communication between computers became a significant problem. At first the maximum performance was on 16 processors, then 12, then 8, then 4 and finally I decided to port it to a UNIX workstation. At this time, RasMol when run would only pop up a picture of crambin, rotate it 20 times, stop and print out how long it had taken. At this point Andrew said wouldn't it be nice to change the colours of the atoms, wouldn't it be nice to display a wireframe model. This went on for another four years, my knowledge of protein structure completely demand driven. I'd just be given a recent paper on hydrogen bonding in proteins and asked to see if RasMol could be made to implement it. By the time, I came to complete my PhD RasMol was already being used in four departments within the University of Edinburgh. As a parting gesture, I made RasMol (UNIX and PC) public domain and placed it on the internet in June 1993. And once again the requests started over... Although RasMol was public domain, version 2.0 included a line to the effect "if you find this software useful, please send $40 to me". However, the first two cheques I received were from a childrens hospital and a neurology department at a genral hospital, yet several multinational corperations had taken copies of RasMol. I sent the cheques back and never cashed them. With the exception of version 2.6 (due to be released soon), a new version of RasMol has been announced every four months. In January 1994, I was offered a temporary (2 month) position at Glaxo Research and Development. Glaxo admitted that it made large use of RasMol, but had no mechanism for paying for RasMol accept giving me money to work on RasMol there. Until then I'd been scraping a living a a research assistant/lab demonstrator in Edinburgh. This "temporary" arrangement has lasted a year and a half, allowing me access to a (bio)chemistry oriented user-group and provided the funding for me to develop a Apple Mac version at Glaxo Research Inc in the US. RasMol is probably one of the most widely used molecular graphics programs in the world, currently being used at over 10,000 sites. For example, the French Education Ministry will shortly distribute RasMol to every univeristy, secondary school and teacher training college in France. In addition to both English and French, the RasMol user manual has also been translated into Thai. RasMol is very heavily used in under-developed countries were universities can't afford the high prices of commercial software packages. I hope this provides some insight. You might want to display a short extract from one of RasMol's ChangeLog files that lists the continual evolution of RasMol thanks to bug reports, suggestions and improvements provided by the community. As you now realise, I have had no (very little) formal biochemistry or molecular biology training, and much of what I know has been a result of conversations and e-mails with people much brighter than I. For example, Andrew Coulson was been very patient in explaining why handedness was important. In computer graphics, it doesn't matter whether the z-axis goes into or out from the screen. Similarly, I once knocked on his door and explained that I couldn't find the second chain in a PDB file of a tRNA when I was trying to determine nucleic acid hydrogen bonding. It was even Stryer (of text book fame) who first suggested a pause command. Some of the improvements in RasMol were suggested by David Richardson, Mage's author, while I was over in North Carolina. Hence the real credit for RasMol is from the user community that have provided the feedback on how it can be further improved. My single contribution is a little used menu item on the options menu. Roger -- Roger Sayle INTERNET: ras32425@ggr.co.uk Glaxo Research & Development (GRD) rasmol@ggr.co.uk Greenford Road, Greenford Tel: (+44) 081 966 3567 (direct line) Middlesex UB6 0HE, UK. Fax: (+44) 081 966 4476