Libi_Bild_Post_final

Citizen Scientists Wanted! In the Fight Against the Coronavirus

Est. reading time: 7 minutes

“I asked myself how I, as a computer scientist, could contribute in a meaningful way to research on the pandemic. […] I have little time, but some unused computing capacity. In light of the severity of the situation, I am willing to bear the temporary significant increase in electricity costs if it means that researchers can work on a drug faster.”[1]

While precautionary measures are taken worldwide and solidarity networks are emerging locally, scientists are working day and night to develop medication and especially a vaccine against COVID-19, the new coronavirus that is spreading globally. But this needs time. In order to accelerate the processes some projects reach out to the public for help. By participating in citizen science projects volunteers can now contribute to the research on the virus, both with their resources and with their creativity. Increasing numbers of participants in such projects show that this meets the desire of many people to do more – besides staying at home and washing their hands – to stop the rapid spreading of the virus.

Online citizen science describes the practice of involving the general public (a.k.a. “the crowd”) into scientific projects. The aim is to solve a specific scientific problem that scientists and their computers alone are unable to solve at all or in foreseeable time. Oftentimes, the tasks to be solved are time consuming and include a large amount of data to be collected and/or analyzed.

One of the unsolved questions scientists around the world currently need the help of the crowd with concerns the protein structure of the coronavirus. The coronavirus detects and infects human cells with the spike proteins on its surface (more information on this). Knowing more about these structures and creating proteins that can bind to the spike protein of the coronavirus to blog it is essential. One general problem in protein structure prediction is that there are numerous ways that a protein may fold. This is where the crowd comes in.

Three citizen science projects that try to tackle this problem with the help of the crowd and that have gained increased attention over the last weeks are Folding@home, Rosetta@Home and Foldit ­– and each does it in their own way.

Folding@home is a distributed computing project for simulating protein dynamics based at the Washington University School of Medicine in St. Louis, Missouri.[2] After downloading the software, users help advance research on cures for diseases simply by running the software while they’re not working at their computer, using idle CPU-cycles of their computer. With the announcement of Folding@home in late February to specifically focus on coronavirus, the project has received huge response and many new users who want to help fight COVID-19. According to Greg Bowman, the project leader, about 400,000 new “folders” have joined the community (Strait 2020). With all these processing powers of the provided personal computers, Folding@home currently is the largest supercomputer in the world. On 25 March the community crossed the exaFLOP threshold with more than 1,000,000,000,000,000,000 operations per second.

Libi_Blog_BildimText

The more people download Folding@home and provide their computers for running simulations, the better the “chances of hitting the jackpot”, i.e. finding a viable protein structure.

Rosetta@Home, like Folding@Home, is a distributed computing project that started around 2005 where volunteers speed up the design of new proteins and the prediction of their shapes by running the software. It was developed at the University of Washington in the Baker Lab, which is part of the Institute for Protein Design. It has recently also devoted itself besides other projects to research on the coronavirus. Besides driving progress in protein prediction directly, Rosetta@Home also contributed in an unexpected way: it served as inspiration for a totally new mode of citizen science in protein research ­– it spawned Foldit.

In contrast to the previously described projects, the approach Foldit takes is based on harnessing human creativity. Its origin story is as follows: The Rosetta@Home software comes with a screensaver so that the volunteers who contribute their idle CPU-cycles can actually watch their computer work on protein folding. In the mid 2000s, some of them not only watched this process, but also realized they could spot inefficiencies in the way the software was folding the proteins (Markoff 2010). Brian Koepnick, a Foldit scientist at the Baker Lab, explains they received “a couple of requests to be able to interact with the computer. So that kind of, I think the legend goes, blossomed into an idea to make an actual game”[3]. So, the Baker Lab and biochemistry department, together with the Center for Games Science at the University of Washington, developed Foldit, a multiplayer puzzle video game about protein folding and design where participants solve complex biochemistry puzzles. Participants are asked to fold the structures of proteins as efficiently as possible. Players don’t have to be scientists or expert in biochemistry, Brian Koepnick explains, on the contrary, it sometimes might even help thinking out of the box when you’re a non-scientist. Seth Cooper from Foldit describes that the main idea behind Foldit is to

”let people help through biochemists and their research in protein structure mostly, by trying to apply human spatial reasoning, and creativity, and problem solving where purely computational methods would [..] not work as well. So it’s kind of the humans and computers working together to try to solve challenging problems that neither humans or computers would be able to solve individually.“[4]

The player’s solutions can then be used by scientists to advance research on various diseases like HIV/ Aids and Cancer or like in these days to discover new antiviral drugs against the coronavirus. Lately, the University of Washington’s Institute for Protein Design asked Foldit players with new puzzles to create a protein structure that can bind to the “spike protein” of the coronavirus.

In each project, individual participants can contribute to science and research on the coronavirus right from their home. However, a key difference between Folding@Home and Rosetta@Home vs. Foldit lies in the form of contribution. Whereas Folding@home and Rosetta@Home rely on volunteers providing the spare cycles of their computers, also known as voluntary distributed computing, Foldit uses the “spare cycles” of humans. This kind of approach is also known as human computation since humans help out computational systems by taking over tasks which algorithms cannot easily solve on their own. Analogous to sparing computing resources in Folding@Home and Rosetta@Home, the more of their leisure time people spare for playing Foldit, the more they increase the project’s chances of helping develop antiviral drugs for fighting the coronavirus.

In my doctoral project I investigate these new forms of human-software collaboration in human computation. I’d like to better understand the different role allocations within these systems and the meaning this approach might have to individuals who would like to dedicate their free time or computing power to science. Furthermore, with increasing numbers of such collaborations of scientists and volunteers, the question arises of how this impacts scientific fields in the way we know and understand them. For example, could the current joint fight against the coronavirus, which shows the huge interest in collaborative science, push the empowerment of citizen scientists and lead to democratization processes in science?

 

 

[1] Participant of Folding@Home (translated by the author). Data collected by the author using a written questionnaire.

[2] It is a collaboration project of various scientific institutions: https://foldingathome.org/about/the-foldinghome-consortium/.

[3] Brian Koepnick 2020: Interview with the author on 22 January 2020.

[4] Seth Cooper 2020: Interview with the author on 31 January 2020.

 

Sources:

Bowman, Greg (2020): Coronavirus – What We’re Doing And How You Can Help In Simple Terms. In: Folding@Home, 15 March 2020. URL: https://foldingathome.org/2020/03/15/coronavirus-what-were-doing-and-how-you-can-help-in-simple-terms/ (Accessed 27 March 2020).

Folding@Home (2020): https://foldingathome.org/about/the-foldinghome-consortium/ (Accessed 27 March 2020).

Folding@Home (2020): Thanks to our AMAZING community, we’ve crossed the exaFLOP barrier! That’s over a 1,000,000,000,000,000,000 operations per second, making us ~10x faster than the IBM Summit! [Tweet, 25 March 2020]. URL: https://twitter.com/foldingathome/status/1242918035788365830 (Accessed 27 March 2020).

Haydon, Ian (2020): This COMPUTER GAME could help stop coronavirus. Online video clip. In: YouTube, 3 March 2020. URL: https://www.youtube.com/watch?v=gGvlNo3nMfw (Accessed 27 March 2020).

Markoff, John (2010): In a Video Game, Tackling the Complexities of Protein Folding. In: The New York Times, 4 August 2010. URL: https://www.nytimes.com/2010/08/05/science/05protein.html (Accessed 27 March 2020).

Strait, Julia Evangelou (2020): Crowdsourced supercomputing project sets sights on coronavirus. In:  Washington University School of Medicine in St. Louis. URL: https://medicine.wustl.edu/news/crowdsourced-supercomputing-project-sets-sights-on-coronavirus/ (Accessed 27 March 2020).

University of Washington (2020): https://boinc.bakerlab.org (Accessed 27 March 2020).

 

mm
Doctoral candidate and research assistant at the Institute of European Ethnology and Cultural Analysis, Ludwig-Maximilians-University, Munich. Her PhD project is about human computation systems (HCS) in citizen science projects. The research focuses on the human-software interrelations and the impacts of HCS on our understanding of the daily spheres of play, work, and science.


Facebooktwittergoogle_plusmail

Did you enjoy reading this? Share it with your social network.

Submit a Comment

Your email address will not be published. Required fields are marked *