Skip to main content Skip to footer

Computing Power Helps Researchers Unlock DNA’s Mysteries

Case study: Computing Power helps researchers unlock DNAs mysteries

oxDNA, a novel computer model developed at the University of Oxford, is shedding important new light on how DNA behaves – underpinning valuable breakthroughs in fields such as improved drug delivery within the human body.

Working at the interface between chemistry, physics and biology, a multidisciplinary team led by Professor Jonathan Doye and Professor Ard Louis has established oxDNA as the world’s premier modelling & simulation tool of its kind.

Key to success has been the High Performance Computing capability of both the university’s own Advanced Research Computing (ARC) resource and the Emerald supercomputer funded by the Engineering and Physical Sciences Research Council (EPSRC); harnessing these, the team has achieved a step change in the size of DNA system that oxDNA can study.

Deoxyribonucleic acid (DNA) and its famous double-helix molecules are a key building block of life. Ever since its discovery, scientists have used increasingly sophisticated experiments to probe DNA’s innermost secrets. But such experiments only measure behaviour. To understand that behaviour demands the ability to visualise what is happening at an underlying structural level – which is where modelling & simulation enter the equation.

“Almost a decade of largely EPSRC-funded development work has propelled us to the point where our model is delivering unprecedented insights into why DNA behaves as it does,” Jonathan Doye explains. oxDNA is a so-called ‘coarse-grained’ model that operates not at the atomic level but at the slightly larger scale of nucleotides – the organic molecules that form the basic subunits of DNA. It generates phenomenally accurate descriptions of the fundamental physical processes that shape DNA, such as its ability to self-assemble.

A crucial role is played by the huge computing power delivered by clusters of Graphics Processing Units (GPUs) at the ARC and at Emerald (a now decommissioned supercomputer housed at Rutherford Appleton Laboratory in Oxfordshire and developed by a collaboration involving Bristol, Oxford and Southampton Universities and University College London). “Developing an improved version of oxDNA so it can run on these GPUs has provided the gateway to studying much bigger systems over long timeframes as well as producing results quickly and effectively,” Ard Louis comments.

The value and versatility of oxDNA are already being demonstrated in two distinct fields:

  • In DNA biophysics, the model reveals what happens when DNA molecules are pulled, stretched and twisted, and how DNA behaves when strands come together to form the double-helix arrangement. For example, the team has revealed the processes that occur when undertwisted DNA wraps around itself to form plectonemes – similar to when a twisted rope wraps around itself. oxDNA has shown that it is favourable for DNA’s double-helical structure to break down at the tips of these plectonemes because the resulting ‘bubbles’ of broken base pairs allow the DNA to more easily bend back on itself. “With oxDNA, we can actually watch it all happen,” says Doye.
  • In DNA nanotechnology, the model makes it possible to visualise in detail large DNA nanostructures comprising tens of thousands of nucleotides. In particular, the model has been enhanced so that it can accurately predict detailed structural features – such as how these large nanostructures fluctuate in shape and how they bend and twist in response to internal stresses that have been programmed into the structures. This kind of understanding is vital to the development of biocompatible nanotechnology able to function inside living cells.

oxDNA is now free to download. In view of the expense involved in coding a model of this kind, such ready availability will only promote its use by researchers worldwide.

Its unique capabilities are already being widely exploited. To take one example, in the US and Israel tiny, drug-bearing DNA capsules are being developed that are held shut by DNA programmed to allow the capsule to open only when it comes into contact with specific types of target cell. OxDNA is being used to visualise the opening process of these capsules. One outcome could be huge benefits for cancer therapy by enabling drugs to attack diseased cells without damaging healthy ones.

Developing an improved version of OxDNA so it can run on GPUs has provided the gateway to studying much bigger systems over long timeframes as well as to producing results quickly and effectively

– Ard Louis, Professor of Theoretical Physics, University of Oxford

“oxDNA offers enormous scope to explore the realm of the possible,” states Louis. “Think of the potential to develop sensors operating inside cells to gather and transmit information. These are exciting times for DNA research and oxDNA is now a pivotal part of the toolkit.”

And the next big challenge? “We can already use oxDNA to see how small things assemble and to see the structure of big things,” Doye concludes. “The next step is to see how big things assemble and explain exactly why problems can occur.” Through a combination of leading-edge modelling & simulation and world-class HPC capacity, this is a goal that the team is already hard at work pursuing.  

Images from left: 1 & 2: Plectonemes in supercoiled DNA that exhibit bubbles at their tips (green); 3 & 4. DNA nanobot drug-delivey capsule in its closed and open forms, 4. 

Project details


Profile picture - Jonathan Doye

Professor Jonathan Doye

Physical & Theoretical Chemistry Laboratory, University of Oxford

Profile picture - Ard Louis

Professor Ard Louis

Rudolf Peierls Centre for Theoretical Physics, University of Oxford