Bristol, CNRS and Oracle join forces against tropical disease

Scientists at the University of Bristol, CNRS in Grenoble and Oracle have teamed up to identify a candidate synthetic vaccine against a mosquito-borne disease, Chikungunya.

There is no vaccine for this deadly disease, which has appeared in France and Italy, as well as in its tropical heartlands.

Phil Bates, a cloud innovation architect at Oracle, and honorary professor of computer science at the University of Bristol, recounts how Imre Berger, director of the Max Planck-Bristol Centre for Minimal Biology in Bristol, came to the supplier in mid-2018 with a mountain of data the team wanted to analyse to identify a vaccine candidate for the virus.

They had the data from a cryo-electron microscope, but it was “taking ages produce accurate models”, which was requiring “an enormous amount of compute”.

However, the system that the Oracle and university science teams produced was about more than just big compute – it was about flexing the infrastructure to be apt for each stage in the science project’s workflow.

“The elasticity of platform we built means we had optimal hardware at each stage of the process. Essentially the hardware required was composed on the fly, enabling workflows to be executed quickly and at a low cost,” said Bates.

An international team of scientists at the University of Bristol and the French National Centre for Scientific Research (CNRS) in Grenoble worked on the project, which has issued a vaccine that can be stored at warmer temperatures, removing the need for
refrigeration.

The project’s findings, recently published in the journal Science Advances, reveal results for a Chikungunya vaccine candidate, which has been engineered using a synthetic “protein scaffold”.

Chikungunya, a virus transmitted by the bite of an infected mosquito, typically causes crippling headache, vomiting, swelling of limbs and can lead to death. Even if a fever ends abruptly, chronic symptoms such as intense joint pain, insomnia and extreme prostration remain.

Formerly confined to sub-Saharan Africa, Chikungunya has recently spread worldwide as its mosquito host leaves its natural habitat due to deforestation and climate change. 

Pascal Fender, virologist at CNRS said in a press statement: “We were working with a protein that forms a multimeric particle resembling a virus, but is completely safe because it has no genetic material inside. By chance, we discovered that this particle was incredibly stable even after months without refrigeration.”

Max Planck-Bristol Centre’s Berger said: “This particle has a very flexible, exposed surface that can be easily engineered. We figured that we could insert small, harmless bits of Chikungunya to generate a virus-like mimic we could potentially use as a vaccine.”

To validate their design, the scientists employed a technique recently installed in Bristol’s microscopy facility headed by Christiane Schaffitzel, co-author of the study.

Cryo-EM yields very large data sets from which the structure of a sample can be determined at near atomic resolution, requiring massive parallel computing.

Using Oracle’s cloud infrastructure, the research team developed a novel computational approach to create “an accurate digital model of the synthetic vaccine”.

University of Bristol IT professionals Christopher Woods and Matt Williams, together with developers at Oracle, implemented software packages on the cloud.

“We were able to process the large data sets obtained by the microscope on the cloud in a fraction of the time and at much lower cost than previously thought possible,” said Woods.

The project had free access to Oracle’s Gen 2 cloud infrastructure, said Oracle’s Bates. “Usually, such researchers will be sharing on-premise high performance computing resources with other scientists. Here it was about optimising workflow, scaling up the resource, then moving on,” he said.

Although a co-author on the Science Advances paper which describes the experiment, Bates is a physicist and computer scientist by background, not a biologist.

Nevertheless, he believes this style of computing opens up a “new frontier for synthetic biology, building simulations that accurately model how biology works – essentially digitised biology. For example, when there is a new flu virus, you can respond to that more swiftly.”

As a supplier, Oracle benefits from such collaborations, said Bates, in “downstream applications to pharma, for example”.

“We learn to make changes in the cloud platform in terms of ease of use. For instance, in this project we developed the ability, for the scientists, to be able to programmatically understand the amount of compute they would need at each stage of the workflow,” he added.