Hartree Centre supercomputers to speed up potential Coronavirus treatment identification

The Hartree Centre team is working with the Folding@home project to help simulate potential treatments for patients who have contracted COVID-19.

The Science and Technology Facilities Council (STFC) Hartree Centre is providing supercomputing power to assist in global computational drug discovery efforts to help tackle SARS-CoV-2 (the Coronavirus which causes COVID-19). The team is working closely with Washington University School of Medicine​ who lead the Folding@home project, which allows a global community of contributors to lend unused background capacity on their personal computers to power simulations of protein behaviour. The simulations then aid in the understanding of diseases and the discovery of potential treatments or therapeutics.

While there is plenty of compute power available to run the simulations of drug/protein interaction across the distributed network of personal and academic computers, the Folding@home team faced a bottleneck in designing the drugs to simulate, which requires a larger initial resource of computing power to carry out complex and memory-intensive statistical sampling methods. 

This is where the Hartree Centre’s supercomputing capability comes in – running adaptive sampling methods at scale across the 800+ KNLs on the Hartree Centre’s Scafell Pike supercomputer (an Atos Bull Sequana X1000) removes the bottleneck in the generation of different drug structures, enabling the simulations of their interaction with COVID-19 proteins to be distributed across the network of thousands of Folding@home users faster and use the network to its maximum capacity. This means the team can get the results faster and more efficiently.

We have a hugely powerful supercomputing capability at our disposal here at the Hartree Centre so our staff were naturally looking for opportunities to contribute to global computational efforts to tackle the COVID-19 pandemic. The way this project works is to take a possible compound and use computer simulations to see how it interacts with the virus. It’s not a way to provide a vaccine, but if suitable antiviral compounds are identified, it could help to treat patients who have contracted the virus, which could help them to get better more quickly and reduce the burden on critical healthcare services.

Alison Kennedy, Director of the STFC Hartree Centre

The team hopes to identify antiviral therapeutics that disrupt one or more of the proteins necessary for the lifecycle of COVID-19, which would help to prevent the further spread of the virus.

Dr Maxwell Zimmerman from the Folding@home team, explained more:

“There are many computational tools that help design small molecules that can bind and potentially neutralize a protein. These tools typically rely on the existence of a structural model with a large pocket opening to dock a small molecule. Our goal is to use molecular dynamics simulations, specifically our goal-oriented sampling algorithm, FAST, to identify cryptic pockets that will aid in the development of small molecules targeting the SARS-CoV-2 virus. Our FAST method has been successful in identifying cryptic pocket openings on many protein systems, so we are hopeful that it will be successful in helping us tackle this urgent pandemic.”

​

What is Folding@home?

Folding@home (F@h) is a distributed computing project for simulating protein dynamics such as protein folding and movement, essential in understanding mechanisms of disease. It brings together a community of citizen scientists who volunteer to run simulations of protein dynamics on their personal computers with insights from this work helping scientists to better understand protein interactions, providing new opportunities to develop therapeutics.

F@h released the first wave of projects for simulating protein targets for COVID-19 on 10 March 2020 and will continue to release new simulation projects as more data becomes available. 

Anyone with a personal computer can contribute directly to the project

The Folding@home project is playing an essential role in understanding the mechanisms of disease, and most importantly COVID-19. It is bringing together a community of ‘citizen scientists’ who are volunteering to run simulations of protein dynamics on their personal computers with insights from this work helping scientists to better understand protein interactions, providing new opportunities to develop therapeutics.

Find out more about Folding@home, including what you can do to help at the website.

Find information on how to get started and download the software here.  The Folding@home team recently held a Reddit #AskMeAnything for the public to ask questions about the project. Read it here​.