The longstanding Folding @ Home program to accomplish the hugely complex task of solving molecular interactions has reached an important milestone as thousands of new users log in to get their computers up and running. The network now includes an “exaflop” of computing power: 1,000,000,000,000,000,000 operations per second.
Folding @ Home began about 20 years ago as a way – innovative at the time and pioneering for SETI @ Home, which is now in hibernation – to solve computationally intensive problems and to distribute them to individuals for execution. It is a raw supercomputer that is spread across the globe, and although it is not as effective in computing calculations as a "real" supercomputer, it can solve complex problems in a short amount of time.
The issue in question that this tool (which is managed by a group at Washington University in St. Louis) addresses is protein folding. Proteins are one of the many chemical structures that make our biology work, and they range from small, relatively well-understood molecules to really huge ones.
The thing with proteins is that they change shape depending on the conditions – temperature, pH, presence or absence of other molecules. This change in shape often makes them useful – for example, a kinesin protein changes shape like a pair of legs that takes steps to carry a payload through a cell. Another protein, such as an ion channel, opens to allow charged atoms to pass through only when there is another protein that fits like a key in a lock.
Some of these changes or twists are well documented, but most are far from being completely unknown. However, by robustly simulating the molecules and their environment, we can discover new information about proteins that can lead to important discoveries. For example, what if you could show that another protein, once this ion channel is open, could block or close it this way longer than usual? Finding this type of opportunity is what this type of molecular science is all about.
Unfortunately, it is also extremely computationally intensive. These inter- and intramolecular interactions can endlessly grind supercomputers to cover all possibilities. Twenty years ago, supercomputers were much rarer than they are today. As a result, Folding @ Home began coping with this type of high computer load without buying a $ 500 million cray setup.
The program chugged all the time and probably got a boost when SETI @ Home recommended it as an alternative to its many users. However, the corona virus crisis has made the idea of making one's own resources available for a larger cause very attractive. As a result, the number of users has increased enormously – so much that the servers are having trouble getting problems to all computers to solve them.
The milestone it celebrates is the achievement of an exaflop of processing power, which in my opinion means one sextillion (one billion billion) operations per second. An operation is a logical operation like AND or NOR, and several of them together form mathematical expressions that ultimately add up to useful things like the statement: “At temperatures above 38 degrees Celsius, this protein deforms so that a drug can bind to it can and disable it. "
Exascale computing is the next goal of supercomputers. Intel and Cray are building Exascale computers for National Laboratories, which are expected to go online in the next few years. However, the fastest supercomputers available today operate on a scale of hundreds of petaflops, or about one-half to one-third the speed of an exaflop.
Of course, these two things are not directly comparable – Folding @ Home collects the computing power of an exaflop, but does not work as a single unit that works on a single problem, as the exascale systems do. The Exa label is intended to give an impression of scalability.
Will this type of analysis lead to coronavirus treatments? Maybe later, but almost certainly not in the near future. Proteomics is "basic research" in that it is essentially about understanding the world around us (and in us) better.
COVID-19 (like Parkinson's, Alzheimer's, ALS, and others) is not a single problem, but a large, poorly defined group of unknowns. The proteome and the associated interactions are part of this set. It's not about hitting a magic ball, it's about creating a foundation for understanding so that when evaluating potential solutions, we can choose the right one even 1% faster because we know that this molecule acts in this situation.
As the project stated in a blog post announcing the publication of work related to corona viruses:
This first wave of projects focuses on a better understanding of how these coronaviruses interact with the human ACE2 receptor, which is required for virus entry into human host cells, and how researchers may be able to design them by designing new therapeutic antibodies or small molecules that may interfere with their interaction.
If you want to help, you can download the Folding @ Home client and donate your spare CPU and GPU cycles to the cause.