Screensavers and Proteins Part 1

I can type because there are proteins that help me do it. The proteins in question enable my muscles to contract and slide over one another among other mundane tasks. They do the same for you and all of our proteins deserve enormous respect. And despite enormous research efforts, we don’t yet know enough about proteins.

Back in the era before computer graphics when Evans and Sutherland were making the vector graphics computers for a slightly misjudged future, Apollo came along with raster graphics and the future of graphics workstations was assured. One of my colleagues, MP, used his new skills with the new graphics to deploy a ‘screensaver’ to prevent the screen burns suffered on the phosphor green screens.  MP had a sense of humour and would trigger the screensaver remotely from across the network to scare new users. One morning, he invoked and dispatched a screen melt, that is to say the graphics on the screen went waxy and simply slid down the screen, dripping towards the desk. I was driving a presentation to investors at the time and knew nothing of  screensavers. The bankers were horrified, thinking the screen literally melted in the intensity of deconvolution or timeslicing or whatever I was at. I resumed once MP came to see how the prank had gone wrong.

A few short years later, I would misunderstood the significance of PCs for the future. I was too deeply involved with mainframe and other big computers to see the potential of the PC. Six months later, I had one on my desk and within a year, we had one at home. Lotus 1-2-3 on the 8088. Soon we had upgraded to 80286 and I was writing batch files to personalise the welcome screen . My screen would let me know how many people were born since I last logged on, or died or how much the population increased. That was when we were 5 billion souls enabled by proteins we didn’t understand.

Ah the heady days when we competed for the coolest screensavers. I enjoyed the craze for Mandelbrot screensavers that automatically generated their never ending display of ever changing fractal images in magnificent colours.

By the 90’s we were running ahead of computer power. That’s to say the increase in the scale and complexity of the geophysical data in my day job was exceeding the development of computing technology. We had to get clever and one day, we, in the Perth branch of a global company, did a deal with a university in Melbourne. They needed instant graphical responses for medical research  but couldn’t afford the computer to achieve it. But most academics only work at their desks from 9-5 and Monday to Friday. So we bought the nights and weekends and suddenly we had cost effective supercomputing for 16 hours on weekdays and 24 hours on weekends and holidays. And they had the response times they craved.

That logic, putting the idle computer time to work was behind so called ‘distributed computing’ which is best thought of as crowdsourced calculation on a huge scale. SETI@home started looking for extraterrestrial life in 1999.  The Folding@home project was started by Stanford around 2000, looking at how proteins folded. Protein folding is the process by which a protein structure assumes its functional shape or conformation.

Basically, home computers were being used to make complex computations while in screensaver mode. 30,000 PCs were running Folding@home within months.  That became 200,000 machines with a few years including two of ours.

It was Dr.Vijay Pande and a group of graduate students at Stanford University that first simulated how proteins self-assemble. Computers couldn’t handle the complex calculations efficiently. Proteins typically fold in 10,000 nanoseconds but a single personal computer could simulate only 1 nanosecond of the folding process per day which meant that a complete protein fold would take 30 years to simulate. That was 20 years ago so I’d imagine things are a bit faster after 10 orders of Moores Law.

These days there are teams that compete to donate CPU and GPU cycles to help solve complex problems that cannot be tackled using other methods.  For example, CERN uses a distributed infrastructure of over 170 computing centres in over 40 countries to handle the prodigious volume of data produced by Large Hadron Collider (LHC) experiments.

And of course, screensavers are continuing to use the Folding@home compute power on SARS-CoV-2. There are over 4 million machines, including one of ours, working on this problem today and they’ve already discovered new protein structures.

To download the Folding@home software and start contributing to COVID-19 research, visit https://foldingathome.org/start-folding/.